From b24ef8d018b03a83ed3c8b84c320ee55425a4dd0 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Guillaume=20Redoul=C3=A8s?= <guillaume.redoules@gadz.org>
Date: Sat, 4 Aug 2018 16:28:56 +0200
Subject: [PATCH] Added an article

Displaying all the data from a sql table
---
 .../Installing_bitcoind_on_raspberry_pi.html  |   2 +-
 cryptocurrencies/blockchain_bad.html          |   2 +-
 images/ESREDA/UQ_methodology.png              | Bin 0 -> 127879 bytes
 index.html                                    |  41 ++-
 linux/Reloading_.bashrc.html                  |   2 +-
 linux/Reloading_fstab.html                    |   2 +-
 linux/mount_nfs_share_fstab.html              |   2 +-
 linux/share_nfs_share.html                    |   2 +-
 ...he recommandation engine for articles.html |   2 +-
 pages/about.html                              |   2 +-
 python/Creating_a_sqlite_database.html        |   2 +-
 python/Iterating_over_a_dataframe.html        |   2 +-
 python/Opening_SQLite_database.html           |   2 +-
 python/Opening_file.html                      |   2 +-
 ..._data_from_a_sql_database_with_pandas.html |   2 +-
 ...plotlib_figure_with_a_high_resolution.html |   2 +-
 ...notebook_for_plotting_with_matplotlib.html |   2 +-
 ...ng_data_to_a_sql_database_with_pandas.html |   2 +-
 ...ting_all_python_package_with_anaconda.html |   2 +-
 search.html                                   |   2 +-
 sitemap.xml                                   |   9 +-
 sql/display_table.html                        | 265 ++++++++++++++++++
 tipuesearch_content.json                      |   2 +-
 23 files changed, 331 insertions(+), 22 deletions(-)
 create mode 100644 images/ESREDA/UQ_methodology.png
 create mode 100644 sql/display_table.html

diff --git a/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html b/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html
index 21a3906..4835eea 100644
--- a/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html
+++ b/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html
@@ -286,7 +286,7 @@ maxuploadtarget=5000
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/cryptocurrencies/blockchain_bad.html b/cryptocurrencies/blockchain_bad.html
index 64a7e8b..65f5917 100644
--- a/cryptocurrencies/blockchain_bad.html
+++ b/cryptocurrencies/blockchain_bad.html
@@ -161,7 +161,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/images/ESREDA/UQ_methodology.png b/images/ESREDA/UQ_methodology.png
new file mode 100644
index 0000000000000000000000000000000000000000..9a30c2af98ba800bf014f0a0ef73929483916924
GIT binary patch
literal 127879
zcmeFZbySq^7Y9f<qBPQtBHbWeqadwxmoS8McL^w<bazNM(lew8(xG$?F${?G5JR)C
z-{0<@{eS=3bCz>B%sXe8dEWcnd!KthpU<7o8gCT|aA|NcFfa&|mE^TBFdpb(U|{1u
z!UCRQQq~s&{$P4&DZav}9jDs?E*{vveDe|mqdp$*#sUOfKXz3z^uWL%?EUwPIp9)m
zg@JKjqAdSX$JcBx7vxJeHy7F6*51ygQFLC{u35y}ltQtH^(hValNLMtRbtv2?^Ec*
z$d6(Iub3V`H^6#)kac#FD<mW-Db)Trw80|zwsXSIcQ9Ad=U_Q=1$GF}l)P62KK9=a
zu?RTye~&OM9*{ly-$P=B|NlwA692zJ{<*{d=ba(!lf#&l;^b=F1+TrB_O<`H^RKkj
zP%hiBy!_tg!0PQa5hv=S4S2?ec$3^U;$j`1uxEmbufLTJ_|jzkve)B%di(QS6|{W_
zHpcMx&c^SZUgF|Ti>qv8=(xa&m=PmtbQ=|7t}8H>4qn1j6(=@zMdLG@Dw;_b*=#-V
z7=5lAw`ta21C5><ME;`Ce^x7eLjQRsM)>J9rV}+r0WKB@mtwrQ+bNupE4I8Of(J!_
z-r{yLq@FLz{&OjQp*~S7yeRm6?#G$Hc?H29sf?@3$V^fk<M`h17Lp0-CG*Ya8AlZR
zkX6^w+v6x_e6Sco9NJ4g_o@4BZl>H|i@B=wOJ71~AKn8i+8f#JNsBw_ceX~dW>z;e
z$9omy`OZt_>9LZ#2ET?qIzoO~9DgJ5`vE22ZqVG68ZXo($>Hc!#gVb9ew>fLD2M?j
ze}GHIo%4CcDV%|4-=meLMW7BQ$~ObpGia;TKbY`ts*x1;IaIV%Z0oj~-TIC3=I()$
z_+l6W$6U{;H&39Y`(~Dk&gAm<`A~i;q3^dBg%LrsKB@f655T(en`|}tL8n|LL5Ilr
zVj;+{$2v@j`O}Yp1*%-bnNuwoPxK3>yGp_tes{}>Riew5L{P60;}R52_+eOMq*X5*
ze`X4!`q(wx>n87rs2s3BSOqHBfORzr5TAq`piX@!+_LtKI$j|6z9t^}3;vP#Hga}i
z3{1B}?r-rSbxn*T@#hxep}3pyL>ESMT!o+p6FV{4qYst%Wb6ju&^XHR3*MHOM7v{8
zo`c*eS;UgsBrTVA))zX>jJ-|j)hp@_@o<2be}K;YGzd-=1K!A}bx`D#L)*Fl{O_$W
zbg9`jzJ0)==gT?Hb;1s>1#h|~FWKS7D;j$O|9@$+N=f){JtVu2=s`y9#JHP8WJ<;l
zG0MZhp<y7hKfapExH^VFYA%1IQSb`~x0nYTL+51yxaS!aZXYGKq83q*!rOyqz8&;f
z6SH)4lNND(-Oqz|G%|&1el!jYw~Z8?!c45sgioX+eH^0WKicl%&Y|l6{vF?BsLeS>
zno}*)ri-QDR5`1Pv`VRr9hxyB)Dnl(bM0=IjXjn>x!H?cU!Ya5SL;Nho&F@p`N=x$
z*2bS*oV}t#_mkQ7gB8P?BFHk9r&5J_8jC)dc=JRu4j0dN8Mzr;@s1-COG_f`mSc9?
z^kPov7Atu8*M{TuvIj+$r>h#{gQn`sW7}#2Mg-Xy<$Lvtn;O_=bcgp2MskEZo~2t&
z)HaWwVqO`IH{Io78=t-Ic-^e_cKi0(z*DcpH~}J;-=;Z#^VygV#zssQAZp~})Q7x7
z?6zJ$nqfjEtpeS_Mr}SdZZ%Qv&pCD998M^96wa)X>%T>k&geVNK0+_PwyJx-y4cAv
zk}7Sq@}h2c0cMn6IOl#Lh0{{LgX!jCRGGp&wKvR>C0TKEP{if(#qF&bIzqmanVpg1
zu&<{!=QG2SjPOcSQW=}AMvJicySxujQEQ3xSi$L&!CeU<oT^mKUa{!TfLqH*EdoOQ
zgT4<cy_L+=TX)Y%U8!Zb<Ln6Nqa{(#WJ0A&0-BazDs(AN599m((#hP3#ZUcBXd?d4
zCl5DI%rC!qmlCBsW*fPa5d6VtpBZO(A8aP_C0+?MSHl#@v7ROQ<SqqeZRxOIv<a^t
zzQEip{T8Serx%(Mblk}shUZxq&2g`#8;7NA=a<|pq|Gzzl7(t;tub9wi*<eOIb)wR
z8Qo2>X}p4%b-%y_0w@a;2%sJfCJv7yNJ)@GMNf$P_Dr*XZIC@St+~%N9AWCIZr6Ri
zC)Aa8V0WocX^6&y*<8=+>UYnHY=|qG*iLj=@;ES=RBaFb#b#`&9@pm{SY}M!_54uN
zs5-CEQLx=QaTVtwL2;)yJci^6-F2b8t3CzN-r*=Fy!@#)GrRa@mQ}|iKkvG9*kGhA
zn>tT+WpS!q{)K=KF^9&sawgfeAA&PYxqe@}`vu?FJ8?G~f>y;&mburfg#nIo43Mh&
z7r8>U$vH$hJ=bE*6dgV!a6wUP*Y;Gx@_nAlG^c5<gMB|E5*w*V?*+Et!tG60)*Hh_
zO67X}`TptOa}A1}`wNC^@Py5br7#EQ_oCtidsAInaYdfThQV}rMjW;U9Px4}Zl2vY
z^xPlEN1vY8yRg@gtrg7ocZ>%met-`uSkRs~<*mwW<4klt@w0arC%T&f@=bie7i3g|
z1M<anv3jxoO1e5xjarb`(jX-ro^imr%d9)j4^yhcjN4AnLAC6z34!*^fL&Z?%<GRt
z_&;aeUCb+L>>?zM135X%TI*k?2-&{my1e~CUmyVUAGUonGO_Z0qSouCE;ruh{`$<5
z!*(-zys4}0IW$dik=dLmrOf>DS<kKNbzeNz%!ylV%wFzNg|#pwy3%3LJbHqzg-JyB
zIzzu^h6)AsaKjn--5MKYoaS!}76>$GBI*iSZqK{T20XQ&;<BHFQ|L~5#S5F0I3!uj
zCdl?KPv>oRW^Ag6XT|On^DNnt*;}v6SK9j1vq;2-5y|VgaHCns&t}VxQE`haM>vG`
z3M$lvw{Tw1HXk8V;j$=m7c}uyyR3Ty$e-;AvMrNhvq7*qk-ye{=k9Yfs0Z)Ae8AmY
zXbkN>xRLpcUhMBe8VSVZlq?PhyM{|r-13_L#H=MjtMtCTp6s+S1NVN!!qgc1fCY+P
zWB+AdC;`V0^x9--M2$b_9uZQMe{%qCAHIkk^YCqjXd2hYeb`l*UaJaoI<2ZbtG{Dq
z?hJwVd{s|{Nw@?Z71fsX-Y6#LEI{O4I>h&Xzp;wI=&f<Kk4pH4pFzsAuJ-q<`9~?^
zv=M2W+k-}l2!VOgD>x$RwzvD<vyXb6dx{DQHI_xtdRaO^+9*hx=T%|bWi_L>b3f_1
zW$mtUxDwPt26BVVEeS45mKh3)5X1R)dm-}=r}f@`W0vjiyY-c+zWQ=LC~7bt<8hD~
z7LWKdw@jM^qh-5e60%H9w1s@>2D#163aDT5$iCyU`%0H$M{v7Zuwf#h<NpqxbG|S}
zFW^)Xgr(mck>frx@Q}ekX{RPoYq&;1o+quh!K`EOA<JJgJuVI|ccRgZgSxs&rpbz1
z^GxA&3H&=oAGKST26yg@=BDmL(shH2Y>{UF52R*xT6&guIjm<IHpzT^@hzCxj@)q>
z-ot3Qg)=%-Movzs|9YBtLa>OwZd-h)N_06g<2cr>Pp!H)X@Ks$yp6lp=zyu$d?l^M
z_!TEY*fs=|bT}k9F`qRVA?Qs~(7QjLZ@2x%Lv*DfjV93K;%Ll~?x-~!cOz!kk@iAq
zO>Vh04Prw&`*>+>;7tA+>cgFT_l2Ff?paxF;NujxdJS3sWwM3DJ0B=Wqo67HTsfb{
zh2ZGZB6oD~++jTb2Jc&*I^~%)mGkzwOjP`RJckAoMkTdI!8^^zxK@bhFQ*qlRTr-b
zu1_QKazTq%W@Pu5E2|P%uL-E5<{Q+Er=lTliq>a>ZHn-31pf1@VcfXId!L#iW1YI!
zIft<eSP`}0<ya5bUC>^^HONKt7+0I}<UP3(q5Nfbboq|jg(_oOMFAGfhhn}EF2E|W
z)rvpA<ni1)nW0iNJ4j2q$SpPeb!hPv@w%vp^%t7+&y1?(WgWG01Oc9q33%+HNXL;G
zZ9amN4+=}VYdPKK#<ij?&f`SpdUZIs+h$SQ&OT~1NTVkarfbZ)5`}KbA|10pY6=iF
z>E|IJ_tERiD9eebyxX=fDL_d@dN%o5Vd|RK(^qK>Fcz_Rhcmi+^Rayoql^|^1tn>0
z!-UtNVoM~~x#KMNNpE(g_1Qp4>z~{w45*Ec;h!se?~0<PNV?v^R^k@I=0oC+VxL>y
zw_Pc)3dg++6_Yg&SGfjFlHdOch!-kJM^Z9|fVAGSIqRB#5q@Sbj*v8#k-5?3hWthb
zTb-`Gfkr>!4zK@zY(_B-<v|9T6rHeFWZq&z>xj*mlD|DRxMG<WJ)AJG4;(;|T3a_v
zvT3vw%7@;wK6xv<O{tezO_n*6^ruJ?Nais?opb)|_erF6ookJ;6`f)OyT<pmp!5Nv
zq&x4vrpBRNvA|NcrKH~<yy<G2@uY@3cYu1MzpKHDR$GyQ`imV~jlCGL{n0bkkXv8}
ziGaKfq-rj-qO^_ai*@W3Ss2lN)Gl{B<RE(KnYqtsy1)A%1e+nA`J4zc*C+C~1RVaj
zee*4+oVdICyPy8A^gyVGqD&4k%PfNHawDl`JEBT1qap<@?s*T!i*EAvcV9sD#=gbA
z$Y<;L4CAq~g~2Z9h|f?w&FVLUf+Volp{Kt_5^Gy@<sG1Du_A;IU_LdP{+nI!;=P}0
zGNvKh4B4kGV=UjE26$Ib>gj2y_)bkGA|=EMvt6GqGc6vYcE1@B<}A<IimOJ-{Oo7>
zq;|PHnb)#DF%(-J@is#@w4<ji6Jo(Nu(#E+9$Z8KmqjKvg1S{0x=9}6gtU!KaEl8?
z9#Mt~#OqB^)s{(QKnz}Ni2D_c1S;kLYx|~t`KrsxMihM*c(Ogn^n_$sd9uUv14nj?
zznbJZ%l$f%Y29Q}ky9v|3*t{PDAOETCkVfOjLn|##J!HPwL37o(?J@-Y!7v3qg9lZ
zr@c^}2+X)RIRMfQBv^Hc|72cr39%7(2jN5ZhB?u}HG4INd|p@!=T8n4fZfrh$mmz%
zvvIFG_`2s6x-BoC5Tgx*r+l}=y0LbiPDC7+WsFc;^F-E#*ZvSg%q0r7be6DoNKO_9
z#%A25-X9u{L*LHk4;6qhbohV*yU?9%P-G2rd*eJ$`qQMYyCjsWH&=Cp9t(wNTPGno
z&>zaZiojn!^K>!f)yL!QL%OD04fCi!Ox<|<^|YT1n#?Ksl&UQ}(8>GcXG9&v@LyEG
zWqJBpNlx<FsF)wv5VzO_VXW%HT$<}R?xvKRW0Y4Y@IH0ft%=3hv{<H%Qx3dLAbc}g
z{lYYKlJ)IE(OEQ-(*ICvi2p9A6_y=<4+Q;jheL%RcId40&a8PIv%S)x#!rOSF1ODE
zYSg@~FnQ;+x1uxTgi!2{AJ#ISZ6}`dA5!-RT&|RKq#B`C#CD2ZXu5A6tK*aFh&7`j
z(?k<e;dqfJe@2+%PpKqIDbgDiyPm47NwY7nHml!F&$}f~IU3itG`s?Xw6=a}tNZc}
z9IWNV!en{k@u{6Xo1!9@1qLfwZPR^h4S7bHg=+T%`UkplhoocktqHF;C#*-ut`jZD
zw>uk84aH-C>K3LTE{)Ha!;wF7V)NJUpkag$uT*@%{z#Mmy#9tf3jMR$F{KG+6f1m!
zd?_aPP#F5Tk2gi?fPf=ymCuT}SFCkld}cWEr@_S4O)faD*GM`&Uie-`tCwVIws7?h
zu+Qa_!-DcKy;v&;U1s~M2mwDi=d!)HOEgxdFW6h}@|(_{qb%W{RDt|2<1sIX4>C9b
zyFl}K5K<k8oGELc{u>MN)y-LxUOB)uc>C4!oz}v~M|+H>j?-_tEd8*i_{DlDKNZoF
zk|IUoN5nVUhhI)}Wx`kQMM8^ondT~Y^wJs5zBuf^+Lx3cf6y*elaD+ZVI)P|P-BuS
z!3F{n?>M$Z+(uojVMjw38Lhh3kKT_uC=@~51o!H~yES=m;<5*QHK>60a0VY}BFpK3
z<ZaKYo_OVt&9FlkAN(_8fP<b%E}kFBh_L>A&(|y8@3ZE+jJodBhd46lLAKW|n~&3~
zN&3adphCsVN)Tx162hIA)~;>F>5n!F+L8VNw{Jyq!r7%et(Tu?f`sj-&8Gx&-b>sv
zd<wTzd8yHW;v1@+U2pbJ839X*b>Q(POqGd=_NVbnnafLp6n%3pU84@vWW^^IgKM%J
zBIbJgq%tqIj=L$x{VA7EBP>DWz2>1%7dAb+_X6m@3u)s7kq&1j2zu>jX{>X){_YXR
z{yEChNKqG;vYu2SIJY~49;>PcthHhA@<^Y_@6_bLLrBMKfKw)`R7$8@fbh6}V!llH
zmt<~WrNM!u`?jKS8)7+8_&hj;p-g>d53)+u(-LTfvc1TkQ2@D@o(UgR2eN;8@@6|?
z>F4cR?Y#Ag3<kCnZ6-Hb!~-e@(fqe**^m-pD;vU9LrzfywZD5cb-#d_tv%xuCFPuU
z9!hXd`~2q|7m~g;C-@Hcxt?91RJbg6w4~ux*Km2d%ZmzswJ$u7Y}DDVq_Kfw&DQr7
zw?k6Ho8#8t>iLgvB6g1VmaJf9s=?cpt|NnT4qHKq<`wv}PocTvz9Irsm)=ddR{8M3
zV*B1aZ<le))n61<M-ki<WSZgDa`&~5%E7>iG&%wGzgT=~{`+BFBZ0rcy(hG_ZZke&
zTJLh~(*)P9V=O~@;irgr@sJq<=k*An>h1hj^$u$?>BGrJf^PD77S7!csZR#wXbP2S
zP{iHqY-T`mPiv2-Lf1BPbJp%D?WnXlPGX|QXP9O8X@kF}MBd%z3A-N}K@R%nTzMI*
zCd~Hhz>{LjAkot&F?5%m;ALvvV2#Z#kU5_zkd-9uJ*(xE?+rr%k3%#1CNiRUa(7Zf
zxMdR<HAA6lU9&XpKZXgsy0(pP4#N=pSLpuXALgM}&xiR5zboZwvL?3J9jwdg#1af7
z42*9lxiLr~x2+B;T+zQX1^B<eJQm7f=1dU!TGd>)bsSX)Bvhc+%4CRkZ=bdwbll67
z17M0xb?goIt-Kt~&|(gSE-Ry~i*#fL?((M%^|Y_O0~R_=$Q_#1y6G>O1|cvX7jtWw
z7s4i&LT9U#H{?p<k#B67mKsXZSzAFy2iwX!P7KqPPfvaZtcSl3beHS;bVEqz!_LFp
zSsS*{&-W}E%MJUAtIc7@H5wyYhwAEhMlKZEJR}hbRdmjKvFyh}y6o3%VzMaZ4iuT9
zSzp5j1XDIM;OyVU4IhqAKdK76S{WWZnya=>k5>sL`z;vcOaC|obpE1Mo0Z`O{-dR`
zz*5Pt-B{wKLFBH4MB5FCX^C7UE@w@f4iBm*&6&@gJ!^6fj6^0wTCfL3`%7{1`@8Xg
zj@|zs<drHw|Kn(zn^Y8!q1D8k2kA>b*eI-M)(A29tM+H>p1<}PIj?n#UR9a_ZY#ck
zov&2fWjB?imLpRhHk|hMmTXX9J3yA;$n8a)!dX=1Q1h~mQzrNLjnYAgK9i(*1VMLh
z75mAXY=*q8s?w4XGfdDkbQY|X*GB%$<;>KQTY9LB+(fU)256)!B_6m1M0>_V(rGCh
zP?Eoqt>lq`+Gr(Mf}!>PrNvm<(*DD&t7t_Xrn$w}6rmPY3SM$P8VYt;<M?0odT-!F
z2VxRaI2ng~f*c9cJsmn*-RIXP{6GO}aAnU5Y6v`d(vI_|zkYm(C$}p&LkLy|Cu(FT
z&d0v9El3)eR!z1`9~->OTInqZr?p#I_a|y~mpNak+AQk#n=iwPT<Am_*%NMX5oi`>
zEzR0>9U@Rt9f@pv!p_&&ydnE_Cq`Q%bBJtSH_i=tiW^V-3xdAbVBzm`zSKwP8n<$d
zgOB*oS~?`(8&N98SS*-eq`^n|15kp43I;dx+hL42Jr<zKX7$Fqh>jP_MuvAli_G8N
z(pda3%^BW*H_87mozwqI=lGx`!5WWdfnl+D7+oC}Otf@RQWRg+l&kD|0!NqPV1p%X
zW<U9_xaE@MM!V4~iNOBV4geos5eB-Yg8~?y^X7XprSR6D<7+0Ux}b3)?}LMrpEEZ(
z{caishHq{|=DjIdj2I*&L_izSO;S_Ve@fZW8FoI-1_xmtqqhRzd{bUxfKarRD19d6
zoxn*Aa|a)}4igCeF~F1?-A#{YG;&13sNHRs8&?h<e*cwxBw5v(Gl9|%VLz*+Xu@{A
z>0c6_q92`6Cc)2<FiNnttM3+c6Dh505{PSPzpKk+%|0}3*ra}|<yT*N77mH_Upm!(
z!D!-wo-PCn&Le%c49KDa_V3=Tt1%j+VZrbaj?8Mk8F6;zpjtN7xUU}EK3ddm;d-Er
z&ZAxKws<QsAl9(pUyb6h#J5`bZw-X@s1?<WRS~M?e8eX<0FbVE%!iq2`{QCou*SlW
zSqY_@cI<HTo@`L?79%7-<%H#MINORLQz!VnhO_je0<a<T=Fp1}Dd*jI2Rx%xuaJU(
zYfk+hM}RfBdBeLcby#YBfb)NE78$70;}kc=aVK|hI-gvD!BV7-*S?3|@SdM*fP{>_
zrW<JqDc->C3)?nXR^E#?{#<vw{5sh_gV)jiV;h!R(Bw630Q{=c72duUD9PfYgnyd-
z-td7Q(@E6FZmi2~3(i9DGRUX&G)<Wa%2-5Cc0adb?`8!OwAHAH!g6DqXm)IZ*g#+-
zN;MIyhLwuqxX(XbuZUG{m@ztJ+I=qr?@}65q~7)jpnJDkjw%j+b7_2+LUR0Sqq;Z5
ztrphlHBwKbGshA2B#Y9o9}a(N0!!w+Shys@C0m;^&BMx!e7~9OZ09G8k07_FYsVV9
z|0xz``%(6+-*_udC6ul)o4WPcu935aaDi{!W$mDKk?zFfVleZqIb-og8MxbsY_Qoo
z6!OwRT-K23n!`nr2SM`>NnHGwm+nnNH$N?Gm4U*Fz@PsXejx^ja#A3<5(4C&pfzVo
z?sVisp*{x&%bIPKlj7gZCZJz8y99gB#7@ZvJBz-1swSb$#CA<x5pb`=!MWKAs)2H_
zAi=RF7qJf*Y2a{+Z~m4LK+ra8n75R3!ivGQx>Ng_E?lywSGtft$;;wGw*}*p#(7ws
zx-D3S(q&ShN`o7H&=>#fZcdG#&#cImY=lI!!o{2HdN{v-)-L$cAELo@o+e24eyDq(
z7QB@yzUfz1q{VcegQx_Xc+?Bt6+Sv8#Xrm%U30kz{s5e~`=MAMylqrHsz2z^2Pwdw
z&|xqY812nZrj%{lVtE0qGYkl(3L(RE(#fH7v#&tAvzi?pL#G+q7eaAo13O$6IiLqT
zIa4sBb+?Jbg|(UN$s}&El)Clb@k;a$!^1G+Y1XwOETo1V?xxA)fXGKs|J>5r+lU~z
z`nWZ-^;@j#V3J~Yf^8^!!szG-y|7%FViQ~2XQ19}=Z{kl-dV}IEv5sZLL3s}qSf~r
zoIe635KxcL_#~w;E=OEldb}cwQosH(JLT&CF##)6Ha_hUKdBLQ4m+rUARtS6Y;qB^
zx6|$OJ;WWeo;@MRy|aPlrFd@KnF91<ZRtROLbp*IKUD8+J^ax?x(Kr8E20qGVdv*R
z_HB+1fL_E95->gap~w~M<a7s!!~={UkAdvxS{o(k?GL3y@|qR$;63nxI%ysP)YlCg
z=fxrJb5iV#?P#uDak{y`u-fIJ;daYFk9bsSQoD2M3@qtdlp}dRg%(BCCNl1^Og?Cz
zjcwmPzYE>5YF~^Y90*c?7H}na;+4{%M`YHq{cPr3DRcHJ@Cm5dBBcdWi@++{9X6+o
z1BI)pj~@oHh2K}bif_mABiv7Uc<A6e?1!>-SAFV3(W%XxBFhZ4wPMRlfys1Knp+nl
zXEUCljra_dyYA{UzsM)~R985~)d|}z%|vfm*b}zDi|jtmnN{ys=59t2$w^bQCom{h
zfyeutOGDGk+uZp+lr&>U_~pgBf8$KZpZ_z+Ef^H<Ch-8EQKlxeNQA9$-yDt>Qd~^h
zph`lC{)Xq@;cgntFG6yy)YsY!&7Y^vo93D<WiAfPSDJJ!V-Ia7bws?ZN@AuF`hiJo
z2Jl6nZu0E?HEjrhy1`mHYY3`0q1hJWf7fp6r_Do3e&Yj5QVuQ^739gi)BQfsZmm?g
z+8LPJLf|(lvmVs-k-dn3A=7Kl_c#y!OHdz0B7T5dtoFdMJZX;){UL~6u#}UN*xY;{
zA$w#US3X2<f`~r_w$~_+qUsS~RFl$r#IridlpfNPmyf%!h5~HHdTCHLPwHZ;QtH5{
zkMY(OqmwpHxcv+(eW*A;(@@_5q_*AOyVs(R`BN=nI$>DK(qo-wIQiSbQ}5^!e`zw^
zZ8XHBQS~-URfd7|vap7-ZDXbHqOy=`#omr$=^1%PiEn%(O=1RVTSxapba0=y;%9_N
zt`WtBL{|*%wh(k{58@M;-2#mSGFBQ>=EJgQS^#EBy22ozJnD_LuRrG+LLGpbeh>iO
zc(iB4X1*Iyr+SbHdO<<1n6vXb6xBbu0aQTghypN58a_2GN`Fy+a@L+w3Ilo=Z9ehB
zh>kUrOw!D{<FxCr&DhMX&_@#oXSK_p4`w`g(uq?MwO~Pys;@5+B~FT%0APxnkGAFs
zZXfyf%o`ksU%mYY6+*R*GG2~9E9gKQcO{ytj?Ct$5St;C!;1koeI||`!3b?CHrY(t
z1bt|h%MSt0gd6+sk)sz0U>(!@flw2!Lhf#`2NhXV&f|~;e4!5~UsI;wPL-bnw(Sp6
z+UIy5&1+Ya0p4n?KQ$1$YZnaR=sok~W!yDuPoV4fS-ArE#-uFTNEL2Pi{|ndAa|FM
zM)+9sx90nf&DxyD)o87Y-%E*upD!ULjz82NZ+cCZGTdu&Bm}x%T=y`Epm>GvX3uCJ
zx%1H5bkw?ur4!C=2cGCN-SJ<UT|mHoSA*O9sBv?bDx+-Aadev7bHABX*qeXYN<abl
zWiKM}ZTxYzG{6M^z^T1?KVm=}GVvwNpew{Sa9_5~-Gof3Z6}Rzc)qpR{7*B}0o*Y7
zaSMU(hpHMg_vP%yi&y1*NhL*QZZ+oB#>9tl5DQvAvrlyjw~T8l5{20S75x`oajy@e
zISxa;)ncRoJ>f!dm7~A#qVLA;;PbE3q`nq`r)X2pe5HG7>}ak1`N|ktzN$l8kA901
zyw~|StX0rbDq<TQw_CjGz9Zf=MKZVWjDc>zJoLsHl1wQ<ZRWZ;O7pcmdgu6kGL~E3
zN9Q!6hK|pZN7KmxV<rL9#{KX-8U;|2r0H0smP|KFMP)2(Gf!GZ6G!H`(I8{apeTJ~
z)=<U4n1(oR-w%kN<B-Ufmo?L?Mkbm3g(X!S-(7LNitSmtB36)UsFr^+ytmc_i~Uzp
z$SvH@pN=eQtmi6tsjGQ&UoFC^{(f-#RoA=2?L$leoHS95OcVb81S2h3IX~mAK3v{i
z9fFXYx-$0ui1<Smg=Z)-Y@$RUV|<~iwK)}i5pSKqR#|%t-^0M|+umHeho!{}8zp$S
zDUPIf;Gg-@j=4WO9#|ODe{sm7?988clX;vp@PD1_&-U2T&`*b5hHeiKc#nM+=^mT1
z*a)XGWCUK)FwEH?4aCRp^=pB%An?|M_13AaTz+~@7RI{FmFm+X-{pMil!teENmbYN
z{Ks^Sjy})_*F(br*mMW)T(o+8gYq(>e%pT<cT?Z^dYaG5_;)Na^b58za>KUkKGb+f
zU{_Kx(DpiqvDojHHK92+uxD|pmK(rFZu(D0{(O(5YwU(tyb}f(dqc+e-tBZkG?(Za
zajS1tjy{}`KzAGxu%5gu(D}8)%BH98T%e9~J40DYY<U+FyHa#t3%+ZuKekPfO^uuP
zcCSSZXzMlU^=3|+GJH7RfOjQb%k_sc-N{l`1%E;IQC&RBPwOgy^K$~Bv2p)`@WBv{
z?I0vBGuDu>L=a*ow4yk$4SN)GNLNJ%v_kz}BCG5r(?V^Z_0(njs?Vw&{G1NlG?@|=
zPTs@-;tp|&+wlmdRuYEWeZPB<fSjtRf8WVX{R}5!zkv}4wBh75Rv##bvW?8vfq(Dc
zUOH76{w<K>LL!{28ks8mjFMRYKz=HnQ*kxm0#O|p*^0on-`TBlYr^f*bIf1-UQBT#
zu+FvC#Fy&9@jYU*_Cc>hLFNbC=f4lluAAzLbErz_PK2XJ5~GXma^n3ev4_5wo!+DY
zDw>Uxz{{h2F85Q1G#9CVRWX1_@gGC2aUmW)+F^Xy%N0mG?8v~AED7~J*~vzt(-r?f
zvZ!4z*KEigU)tvnKS_&KW*^JJC^vYO5kbH8^z`%5jjMl!vFpYq(xv+^NVk3i32xIa
zgB<*7g#vAD8#F+(w&h^;K|#-3agz<an!;rVoA`zzV_#R_ov6~%b#Bcnqz=|Tm6*CR
z>=&3`yDPl!<RbqHfVQ2?t1`!ORe@=4Nq?U9@<o`N-M3>U4*aBX%=9;|h?w0f^>jJn
zEcH9T2+t9MAIgXxmdzVS>C7yT333oqHikq8*+3wx`h_tfi7;Nhw=4rsovh(wfYVfL
z+g!)g@dOPE?hXab>I!k!RZ8*V_Ki8`ABA30`I#q?YHH_ObtkP28@!k|pmt$qqy#o>
zXfCm2bz+P+8;`%c<983|mnEdh^z2{p;7Ef@zi~8)j6G<>bU}Yl_WMTnQ{fatTvnP1
zzy3mSX@1i%d!$_dAN*cRcM^enHP)}qwjUTXo<FlU@TS0h02ao^C9_)QMQ2@}A0JZl
zJx6|Y8bqM-eqjen<GF@U9cwZ1=F~ys{DLVqDgSA9W=GKh<gIOO`Jl^BdxYrepO}H|
z9)M${QxBZhVTlx&T>I2*JM;Q%AXC$Wsi)WA#X_}3L6ZahX1M_VQop2Whx?pw4NWP+
z0tkd|f#RHFN~$j6fofuas`vNg%_yBXtj{8wZ=3EioCuxe0F<eJX?@c~7+!L|3V^vK
zVly(ewIKtVMAjOAh!7FJt2K<&t=fKuPMO;o(kO!Mq>yU17Q@iZ@Wv3*!=jn6n=mef
zk;tO^lzs_-0(h*lQagP!NH7QZViq*><qDBE;9pf)G!{8naH`ZAunD!Fa-6xS(c{~y
zwF`FQ-tFve5D541!@On)`O(ufdxKzl{*N)&L;KZT$KN3uOh^;X7X3AA#XS*0)$D~3
zDgfD`2f#?RqLkourHd&eg4U_}0x$7=$re5U5WGAW3DkrN;!xKG{h##I8p7?EWjIsZ
zK)!r$mjGS6*VyGj$-<uO`>-pIUKV62wb{_F_cqzgu>cQyPSHkStJZ4GPHkqUY>_Qk
zXZ9~Z%WmeD1^>4JRqd*Ro1u)|(d7X7SA*X2cGVvpr{tH_Zcx9G&+}|xY<iz0BHoa$
zqRr&2t`E0@E&ZyiTwZ^4l*D*%i!H{6<{+3Tq$5u)2FENRjoLNDd(<GzbS;B(){aNN
z2oFxF)E^!^q+V_FFd>Vg{ng2%Uut$F<juRBIi0&?!Vn?C0F9ynx~7_~L+>`dH`>AD
z{HvNm1tRF93#$ZWrZM&37aYVzfv!LiUn9BYKE1|S&;Je0iSEgg5lbxx$7oF;6I;q#
zYh>5CdQgZAbh29Awof;&M+aGy?A&<Y0p?FR8w&yTjVGQ^K>+O93miF=AMQ;UcoRz1
zI@QTy>?dxpW}bBcaIwO#sD-7|@?TBM?I|5p#8hP{LL#JV0^vSD7b?jLvP|*dkTpuY
z$9UG4ji?jz6Pt3U{T7yw)j8sX1X%NYqt*N9PqKP~$@bp|3QfE-TY~4c@8lE!r47<T
zr0flx&sL0=u*;Li8|Qz~qa*Hn;oradZ8>$D@4cdy5A5?Y?ec^NdH+x@ejp}0dWA5M
zJaNG5lpGd7|L5xOC_~M5qGMp@X?Is8)KsNr5=glmqPkp{Q&(-XKL^v>M_z92gx5SU
zk%qAy7I`8^=IPHLsN!WvRAa8Yn}~Qn;WaV+nK?VrEThn`cKSvMFWp^qI&uEZ1arKb
zN14h{@BN&9Iv^al;3gO;>nfy4V{^(e?2T15%&d0q{j_H?W{Lw5@y3Y1KQmaY(TTGO
zq!gxQWcEAC^%Bqs=So)0X8dVJ;zgz|o2|)t8E+;>IM??2?~;(`bM;{o6hHTQzf|Cc
zq44M1%!!0pj&YVCXb+3JJD+$vbd^<oA^^}mxpntxC`sON|LK0vKF8<%{Knc~QKaCa
zUU7(70WKw&VUx+OKF%nKcF$3btP;;XHZ@p7K2INdD3#NbD*USEE6pMvDNi}uO@x*B
z@ZW<YJJ)y7?v(+<e!`XHXG<SOhN!sJ&EL-)AdQ6O-c}0Ahyen$t(BXBdwd<#dyq;V
zODL+^cr)s-a(kWq`ub&st>%-F&tLO8O{;8%73=m6xkNBKT?qm8+8v*}n~OP8>1rvZ
zhF#`^28=$hllBt3P!CLK;$7^%7U$2}g!rIZ5tH+Jzj>HjR5!wOkmT_C%hnO&oORAZ
z;#Tje3zNsMabx&1JlGv*q{o(~H7`(YQ|@$i?Zo_+S8Ld(-^aQ=LCE~lgk8Z&PuQbu
z)$T?iReenC-hz(6`58Q2n%0D?OnjqhDr3UJ=Z<>;-7vv&dHbx#rwt|gIq*!`PDI2N
zXzTpsRd_1dHH~#BhK0jqt*^ERQ^gJJhI;?bB01kdf4C3z!=g<<jj@bi!uS=(J&X^W
zAhL4z^+UtEu&t_S36?6Fn@;Z4Zt@Ej^)7TQSK3#0lLC(H!{*}9L?ddw#G63!3UH84
zb(Kd-y9A}PT_~!mjF}cytEI;h`$?6d*LnupM$^BK`qKY8tt7^cQT$>8!x+6PFFpt9
zrQp#O?(JkjxYP@>!S$w^))W<Aa~k&0S14ukYh{VDQ95v{TzY}|LVKYp_bw*SRk2fD
zY9cI!@2IMtF8IntIL^Kby#7Hi>G$TS)=Oo`UL3Qbo|q<*+AuQiXqej}=%&)D$EX+F
zZkQ-_S7S7;koo<Lo6<q}4iWh?FzEqz4T`;0u%D=z2D8eZJCu3N0@Y*uf^wuv|434Z
z_<;Trkh%&Y;%d42Q3N=1fDX!SS;5aV{!)~1gd}B_o$;hgXVYZkM1G>Kwy-tzcy!bA
zuJKt0=R0TRkruI-fuGRAdo2vGE;bP8O3Zx6R3~k4yu`S{5GWoY9rj{p;4LFNy~YxN
z4h-xauIy4?OEW2=xI>TQa_e0fmQsQ+9*}3Q1<X%vh}=3XMM&G_-&Ax^4oYD*eL0hc
zg~HhZ*}pCsPz8OB-ag{JH{`l8x5f+mGRjdV`;=I-@kL=81g@T@oQVuF+%i<BJ26i4
z3#Z0lBb4!a(-=F_Vc(oHqF{Jnk!(mjb|L_EXMG6OCc(g{WCf0vS9tn8+<>YRb}M~O
z6tGE~h7+})EOWMZn!k+ofW)?+2|@4tB*<am*{W;b4R54!knpbpMV{nw7^*bo(&PaK
z+mm?6tj}!b<4s{ng3ZrNms`JA=MmZB;8M<oZe=b&+g&O1-^TzL0Z?vo9JWITB($eK
zECRuZxuuWM)93LnT21MF=J!~lSpS$!{>#ovYi}Vy#pV=*@d`h)-95_gFX3Zq!bv~^
zD7tTjFH_lChlK%-!Ukb6HM!cvrbwQR-4Y4Y;e}}DVf_$-rp62a*iq018)jMkIni~D
zf}q7!i0?c44{IB4L+H^c+gsfMSwPs~`P`G>;u|prhK?!FC4QtB4hS%7Jb$Gib!hVB
z=|p}3B<R%<RQg(F6Iqt%lE(ZBsCguXa7G@W#nhPk648V_9-}UHv{&+m&z_@%_POQ&
z<a4ea6Yzn5$^QMiICQ1iz49>i<&ZQ0un&cwDQ1XMKA??P3*Y7TaxvvS6cOk!QUrJe
z$EHyo=heDVyqMkxl~{r}O}TWDE1%qIt(o7{ajNX>8!@o(3A-;2=T*}Iw&ee3OM<hX
z%-$~Lb>rYV39E$G#B!HMzvB~4{d$ysnyF%e%X^X_SL5p%@JwtBj2TM6ZT|aViUGjG
z|2_Qi=>PST{OjT!*(M_UMA!!#I{Lv)3<pV`@wXpZxOI*lP1lX2=k&MF>UG)_jlFM<
z=-@1edZI7lY`5+GTABV^!XxDZFu1p|>=!`n3Unq0IK-t{eP>uHZd$$dpSKl7@0r!{
z;OPQL!(xQgdesTkXL;$Q$$EeY9bsjxV;aOo2b?E>n9@G_zdHg)0ZpPH3Ngyy#E5pb
zyT@@%tZQ%Oo}-}prtYG%xa};x_d8a)eb#Q|6}hbm+tT;QGl`K}r;(X7dfDaIpIZdH
zPHSAnk^=xt?iU4MGz^Rn#6Yoo0G%CeRKJ3Hg7+t*vNj$={sIDR-2a&lJc9+iK#a6~
z;h}V;&+1pOBHsTdTMFy}HLK5NXy(iKKPx`$zkUM}@qh1FZUhefLC^jx<N@x3f7cq#
z0C5VM$a@BOPEeYak8;*Pi@9;XYAdo$ZzB1W{&EwaA6vL+`<QOIltmYRVtwVH@PSu5
zvU+zx`;Sp;)XAWZ3pT@a`(X<SpV$=6`>I3sUu=WC=Aow?jnX)+4$6w*T~`7V&HHeu
zeAR~@>9{wcQtfl?v%IHi-kYCSfXf(dy8!Q6@d)LnvnR_e)SF_(CGJ;g$o`7&((O$@
zR%nyX<O$JNKdqL^YBz`&P39Is{WpTf68{ztZpOeRBQt@KIzr62_Cva9&J9=H)KfYK
zGf{QYf1Q>JB2}+tnW7&~x5S4`WQxoCkZ0!b)O3qag#Q;tk!gpohO3<+d)hLs1b!yQ
znuc=}ml4eM*|z5OX%o?0wjF>pGO5g^W>Q8tQ2HW@rsVcOTR*+FL*EXIlej9Hxs={G
z*)96qj=<Dmi}AM=^ap?-ol~#QM@a7*cZN7E{bHjR>T<J=uNVa5X2+X@7~Iud23LF<
z7^OHt8-+P;rT}-JA|N1V*HZ5yu_4jcTFXA71rQH_g~YX)_ef=12!cmFGZZ@afb!BI
zh+E%y#7Lwz%<#OzoJysfr;6|9jr4HCoLt^7m96lM#dK@sNz!(=rN29Yd4q3LLxt+w
z0l|a6&2<x>bUR+N^*@tg=+d!kXaF#DQZTGl-7)_%RtOg9v}X+5nl78!3Fv;|PzC_y
z0li(G6D*F}d^frA5^1;XBFK5L@#4Dap<o8YP;<6Ma}jL@un(pq3e+KkKD_ZwvA*M3
z3v|mwvz;U6p>*{j^9r3hi)e=Amu~XY4FJ?iGB?y%CvQ5(I5%u}{+MyT_=gSVN!s`8
z5%)jv1fB7tD2D|lc%2{@@#+G}O7Sy5_1bAB_2sO9kCFa<*cJPqboL{2SG4WDjT1dK
zz}KA<19isIU^-R~Hr=qEfgPRW2r$Fqb+h&g5UQmU-sP`S7wQNC111qx-d`s}iFTrM
zIPS|KhP53WTg+wvsv%ua3B*~0`o*vL%C~TI5nb$m@$kqH!0CDl!6FAqSJANoSMXkn
z=gHdW_K7y7_%1Zy;a98in?^#bBIiy&S#I=xGZ#iU{U??^ow+Z<;t1D4;}Ab|nSUs~
zF`P;4$~~AWFfVcgLmA1%MrE};jc;#I@K-^<cc3A@r#=o8Cnr8esC;3v<G=(+Yp+j1
ze<>UI?pSR;ccCRpg-`d4s%myhp5Yc(j1(x2RDGifa^PSk{eD}V0IE?T(6S|v8HSiq
zP95BT$Sda}hULcXm~(^&FFw0Y;lt`$olTXU@;nEW-|&tUDF!Ejf13{`#{mKc5KiJR
z{n2*<oR5Td<k>-WKutll6c7^at;qLF8b^B^o})KL4=L~$=IJ)|+BEsn&k=c@Pjh+R
zk4c>!GXQB8n1DHdd<c{X)_W9752fD42Zm3Hr-U-YBEJAs2}6XFA#zNnq)Y0Jm4fqk
z+RLP$qSLXQ?IueZywMS|+h&vw$C&6VgQ2}LQlLY3YmS95^#g!);+H@DX|u^W%z3Tx
z0dtdrPe_^#UU3&0lWQ_%dRQLa$M}gk%H@4ioJmc5p*r<?`kVTdeBQbMBIeLOr^|%E
z;psA~P3tKKaeBgZugIW2fu1~i7DL-y?8zOc@MwZX{%M7+Pq})~f4N%24cO<W1y~Dg
zfq#q|!TlC=k(;y4ZihyKuJaMHyQ2&U@o%~#LOa^pd{Tv_?d&g?BEkU1z20SvcRj|5
zt<!eEg0}}4sVbrQlD0XL<;*6oDxQ57nh5|7s*|ZRM9Cvx@6QkGz{4$zBjr(TH624=
zmZ${rgN8dfv`tNa0t$tLZ-V7`l+U#48^H{JwoTz_2!nYA1=Umcc^uZ@J$Ste*Q=Kg
z<?ylT>5^<st79m|2?Ne4BoT(<3_><*fVF0^^~ALE%G&^s7CdK>^!k-5nYc3Q`3R6<
z8%g8<LtL+Qz@aH=e!t%bRT}^z0$><yI?H(<3&0iM1dwS?)qiXL-3&;oolHm}PqFPK
zlLIbr4TV=^W=(WiVsQHuXH8}>&_C)ai~!!;8ov~b?e@=u{uO{48o<W`4zxm>p0HGc
zJc8y{1Ubp&y+Bf<pbkjSr0fB*<29*FIja^7Ifl7P>Z|AOKr#WMo;si^?sa#-IoZ{L
zp}V9>aO);aZ{qj1A0*C6%aPD8*iJps!uYN2a5rmFRCy1*k5ffhA)Sm6$!C>auC8r2
z2k{Hpx7k9lR#;A!imcgg#Yl~VO@Nh)n~R~%dt)NlUfe}3YwCuh4M=O)Z(`nmL(|wp
zo^a;ecKNN3TO|OwD>MVnbLNQVSRf%E6sW4$3GN`7sA?3Z${I&5v<}H!2NZ#hU%J(y
z2|~ov<qaIiGH1<OEZ>LBEq-dztfM+n5(x(oHDtzFuG{p;{eSeVT%Lx)mk#6qe5(8k
zAfo`)<>#Eem2w_E*I}T}?Uy>|$Zv1mwu>eQIvy*yh#Z6M#_H9HxVv^{Me7wb0MbyW
zUcKQhW+N(ezLx8j8xw!Vn2j~m@<p3xDG7c8Osp;BP=}wg`AN0o?UTw1a16w_SltZN
z1&nD~<=X_?TcfsB-oDJUq;dBDGL$Jii4g5>3{dPEvsAGGS$;~$oH{MRW>}j|OBBFk
zNcY&`PmWTU_C%`vDs$#C4+WM~<7DakV>ohv*$15&-^|?7x&58O3kKP0Rfyk6({UDT
zqH{=Ru1!W-JWa+rHfz)&I{Peom~L$~n)tp8kGHJ?+3|0z7o!{~O8(uuo9H408t_FE
zl|eR0vZ{NCsvO)`hpdWhxv_dCTabb>Ve|aIEb@;MZL^08bht%F7K*Z|@^@tW^^4x#
z%r^#=w2Vdv$R+Wh-atfbpWtPS)U*tKe?LX_8tg?P#jyQk>O<19cB{Cm>7STG%gUpu
z?UdKSy<`lnke9oYcU_Uu+_;+=_tAoQVJ$j|xJZhZvdfaVjtKZ~5BI^nK(knG^HA<0
zQNePp*=a0p{>E(De7vM~iqv#IzZULvWx|dfx{Z4H)b}XC{p_9M2KI!Eul3~sN>rVA
zg%`HT3;kT`_ImD%+wiyN|Lz%lh6!X-KuzUrFX%IMOCHre(huiEz%wkPIkiLSJVKgc
z@LhrVNZFAm<MV{2jCahd{3_!DoWDk0ut#HAf7s|T^=716bHDCSRGGvsj}CK~!T0Mk
z=i@AlSXL-s162Qkz7=#{7V7+jpQ*?Pfw7Gscj+LKNaO-7siU?>F#phJi@$$S<VvOG
zBBRGl+IV{=HN2=u{TDJntQfCL75yHmhzMK>OXn6<Zv#VZ$Fml4jaWln+*d1NQxO2#
zI{_#7`wGbBb|!9pCo0h(8HIm$^}h%Qgd(5Z76thL{a=<910%b6=)7D^bjsY$GT%h9
z!7DR)#I>n;Vvb|;gDom@B>?Ma865!xAz!P!<gB~1v?Xp)r`=!6Q{a{Q2INr=Pu|<u
z_QM>Do#uu7-xgsh_ZkhW|NTm2aEUNj?h*T)Gx*tuxH=G&MBN&TzI!dov)wE^UQLoF
zd9a+F=HAxM4jjk_Nlo8vHoS+)Xd-g-KuL*(L;aGEVV_TJ+@5<2$$)>xnU0U-{<r>0
zc8!7p*BD$Ksydz%cTZfMdNB)Jvelngg_G`N0KkS<%^r+9UF-;=v8FYPz*4@f+l>~>
z8ItyLIUcBo76EcP_CsxBMU4WEd|FEC$oLyKoBGB~n{JJ=X}kjPR>gUYfkG=F)SY0w
zPMHXdH}X6Dgy%}R{1<k$O#9ygW*!15QtT<NmA2XMqOQrZjtmrRG+?U)K=Fltytw`T
zn^WWbeY}$_Ujyf_P5&#k#}ql4;ZAb-lO9=Q@3(1y$-~pvbc!Y5{_BM}S7wt1pI?^{
zO(ts*5xZxA#0vTLS)>SKILX9;R}1?TL>1Vfcq5ifGvZ?Se`4q5cI0v+&cB6dcmWcP
zPs5r_$TP7BoGaTv9=^A_X5X}(=~&}^?%vlblyCqGSvs-}Qf0PB!7HL~j`PoBOO!g#
zMk>LHE`KZ`J=8A4b~;QbXsYpa(`z~ay&X;@y`Fx!b-j=^xQW08j5L34z4CH9o;mLI
zh>pc!BN*75Ib6d&g_oR=p`AK#j#E&z+m++lf5|dBxK(ZZf5)#h(02S=+Q`&Ni`yYh
zqz|VOfOnsc^I%H`3$5*lS>Ex>H1N6H9WO=;K@#j7ZP&sN-cbL2hBr%zq9Tt<&tg2R
zkM*ypo;}I0UCfjLo4Hl-^5};G$RSTW8fcj2R?i{wFru*+c2I~j9x$0_=;@jrbc5sH
z#+`Mq`n}9ErC*@ak+T3RammjArG`9#pK9tdb-B|DFqI$`LFT&kUxV?1laP#^9YFYD
z?raMRU4M*eG^aA@cj`x0EES}d+|TaXWK&DQf;%VGtfj5DF+r*kH8-x5WkB@PTD^P8
zHDQ*BY1n3yXQ*T^W_f?Kw_40Q<5rVnZ%-!+a<8GXrFp+G%LRG=Nq1~o9ymyBIDVZL
z1Ok-waFZIDbN-3kchP-zZ1V_BL0;QSgLsLj%>#&xRV|yABi;srkpAu+gF?EgZb2hP
z#dn~MF+(%rzBd+hnmIt3(|UICU$>`21XMWyIi&Our*vAB*ILxICYmMvki>f~>$0W1
ztpuB~LEU#%U>nEQ(x>+GY`tE&6hPhZ(AwqXf2iq|Ia3|MX3MJ)kE<XE#3n1G`|Dt^
zVno<FMY1wqu4NLUKl=#r{C8drmlXq6BR(K&%l<7>6Izke6W^8=`)p)4dlHAUSImY{
zDCRq$h1hfxw_K!-%!ca4{iNsVH6K<@HWPZgZ0mPf5e*vRZ>Qh(RC9QrgY)VplBi@I
zq+y>@0zROx;^upVx;0<=gyHa?<LF@jpLdk!kghxtLo`3VCJ39@AI@cWd9WcPpwC=b
z+E8A~{<1qDcL6f(2h!oj?OU)q>)-%J0s+Q7=;9BQop{cKy7b6U(N3NR=wayV_f28R
zVmZ;09ONQa?~nis(mI06>nt&q{7M+Qdzi7Z+$^PXE&(g+&>l2-6bXzdopJkA)Nh}*
zuL76>mq!$kwr#iG35<pk(Vl%5*%HoFS4Y<~a}wJcbkDmpn%b=f00ME>(o^r6oZ-mu
z86DOy8brxYS}>bUtF{vwb+}&HNxc4xFFDgn;ZP$K^g%}$@+}BuTY_!Sq{iB-nT9ol
z7q)flNM|@H7Xh<kbfW|i7yyG|1G1ckGVXI}gwsDot9cj4?O7N4VIVUxPIip3+4H}<
zEAEubg2x}KS$5%ULJxcoBP0OYuG8K}xoxz=v7rFP$xYO`XXdgW*G@OALmbh$RV#~U
z7J_V60w(&%3L|twsnS<6r>8EYQXkB{3F>@u?{Lm8V}^Pj-=mfR+#W2~u{aIF^P~4H
zrL)VNx<=4|;#&O*qpH<)3aM!9ynjNzQNvm1V87%}7KN7aNjm$*Ws!fEDzoq(v;jT?
zpbaw4H=$75y@hVW;BiqS4Qik;_!@8`h<Nl60f%Kx@rci8NZHn8?E}mf!W$MnWfdme
ztMl9X(}wIWH@V*N-l|pp)DOp(v#z7gHgwN`>>M-2DujuDHa2Uj$9^ZIUOFx0VFj{W
zCTw8}*xaR%Jti6pA#^6<e}wS9n(%)2c;Lq@MHPo)sFl@_ln|6=Jf^3=9rr&!`!^R?
zpy7wx$4;@S#`uo*#NJTaaVcRSH(D}kf~vH4_6Q09Y6}XTj)tG)@;<NfCJPGvN9~L>
z*7X5whsaf`AeYpZu>O{Y-#gZX`$r9P*#Lit#TmNias^(f6X-(cKh$IDxAv(q{S)PE
z>2Ene+=592d=mn+wa)7CZ#@Bxr~ej?o~5&O47dmAm~ZMgeVxGq3q%i&7|Cnx#~%b(
z(cgPF0^ebfDv)8kB(p;b@dtDj#_QNtLTVq!dbW!Ne`6hl{WCVW8T1b)0X6OmbvMU{
zm=jqO3XlRg!mLzwjUO<?DY<oPHHsLfp$6JaX`A$w#=7}tsbOGDJ!sh0=x*rD*9)K@
z1VNdNO=`Jo@XHpHG<JYCpC(_xp?e$xo8>+7-G1oW?=?HDBh=@0o<BR<$+9GXR)mPm
zHE8Dbwm1D$NWo3ME#L}arL@jBG9Nq4nEgdEm1eB)B3>xP8jDjaYOK@i6EMGC_Yo#5
zp?w{-2iyBUID6};xRxbuG$A2EAdq0e0|XC{;1VP_4DL>X5AHB%2*E8uhT!fD1Rb0N
zcb6c8!{F{VZ<BM*z4v@?z4hLD`yW_)_U^7-U0qfE>*}hVKCK8|p}h4na{z_~@<P-S
z{I(_y%+9OuF6+)xz|0nx0w1tt|AggQt#|-H91zj3@KXto<(JmRWZkcLfvM6ToD^i7
zO-AY}W`^8=`(Lzj;;8zB<srO2=BFHNR7Kt3pk#)n6Zx~aQqr5V;I8X0tDZh*U8|q#
zReuQpmSWocyo*%s<e7*vs@Pimz)g3V1IZwXDfXB&gRlMSijEGt`yBJMzs7hjB(Mvt
z@<5=T>4I>x<i0K{WJ2Aa`xnUpta+N5)A}?>st=3C4&69Y22pg(pkl-^D)FW`BUb%}
zrMjr2^hynlls3Ea`r#Vp56o!m)Kl!r`(BxiGTmQe&d2hKdnuv3J7Lx8a^;xfwZLZN
z8@$`;FMNy+0KE7=iRRk4uC#!~*tUwG>ualR*`1ZyuVN8?Q--VnP}j5N2mqokwAQ5*
z5j8@@1`of~&ELL_^Yk{g0vdod_VhXf5qPWI@iSL~z?g>ZqAT0eI&%w7XVS~1RP|!V
zKkI*Ye;GK5c8k3Hg&}bM)iZ!Ai)+YSjz?*{U?$tp=4?@W1DZADSlU#Ev_;!`TX)P5
zq<FPyKNACjz<_fHl~?x3D|&?ilPxA}eDMd<r}}!glzg&ye4BRSQ_r(`xg4*Ck~hi*
zSg%G}iDFDUae;Fn;WE?;7~x($i75RZ4^<OrlYqE3kZ=fiZvy~2KNa}aD}b*&OysC-
z9up0By)cfC7|?rR3sIKBt}(F*Jx*4-efG}Sa{w(>yAKEDv_x#`_9yy}D$cUvAyM9x
z5KF~0*Yv6O2I^gE2dKsQpT`S06<^*dcv{5KuM7aA8vXZg!@k2LfCH%otb|d7_xrAg
z;*a2`*CKak7rP7j*&j^idKSajYEpohj{|}nzv+9(KmJ!<#4VRCUd+8V7Fr9%s5g+B
zeVJ9V!b8tW$X5V=^OrsI0$~3?*QYmVe?hW$?mVLY|9dk}zkUx*&+{#sukaT`tLYps
zmAiN3>4DD?K3-f1Tu0qt`wG7ld0S;Q)R=$g4&Ci$PMUi<lOl?O;e_HonlI13{r)-d
zxX3NOeEZAb8drtWBcQpYv+89u;3g$H05B^y4C0&j?#6Ip0?o;L0-yQlEkN?84gX@{
zZ$5)%-yYpZvtPX3hPs8sfrEuk=)}`0Ws=l5k6MGVZ(DircPsP-)Xb4$FL8o!DdI$c
zW|;qJW~o-O-|a7o*8TrQ(ODC{E88E6fSm*!L!k3JV^8#aX>a71aw%Wca^-4}XEL#-
zKN*O2tA22O#G4&fWy-HKbVyflNs91<)${uoY8<LxEXpOEv%<bx^+Bxcceke^bQZrQ
z9D%%KLMv9W1Kp6kwO5AVFa2{v!~5geK}+PfBw+r-lq~%i3bLJVq|zlq%mRoJ`Rr)<
zVeQqJ==I)_^&K7aUmq-Qx*GD<J7Rdc6Ueiq$VGP>qzDZ@6wG$fjjZQNKWyk2sIZE1
zjx~A&WvVm=B{^UuLmo67Qe}&2Yvg+mm0(p`ND(bn3K@9o633G)M7DS2etWjkkxd9T
zBLl>K@D3m;dHQP@9NE~g6vCaM;>UU}rQ<K5A4bmBIAw^CJy(({wL3Be&JWwd_Glh(
z&bmSSd3)%D+UHwAELrc#=*@DRVVU8HCFKM*VI%UH<jmkU%uVev+A|y<WyyDkEfe%c
zQP65D;%tSCqu}9sbVslFXYX%i;y(jtZXu*yM?bSPDOo{rx%p1i$U}MKRl;bXpj&vq
zfaLR%+AxB!L5G9k#A4k2>Pz(yvTY_xxBZC$r*oRAU+&jh#J|5qb&S+6j4rFx&cBmh
z?yf(6#~g32w{I@0k=FGrhF2{#$mK{FYo9_MmG?ATp|n`dI^!6=C^Ylka|Y7E40;xe
zd%^nbc3AHHGb~~6RbNo-*q>T!TVig@6cP+x@Mx2OAcyO-b;fvW%hBdLc3mfv)5GFR
zc>R!W2B({vqcyJ>p7%^ywjS;~K2B{kP7CL35~Bx}+5|A>#kD;Fgd6PjMLi2ADv7*^
zM#yc>;hxtyxrywl$zh}!`4DE7w175S{CKO>@i;Toy$URHIxC$zmaXJ7(tSRTe|)wq
zS9`!V=@ger?^^7f51dab;cU)$+t@nx_Gm;S!Lnh%=8fp{-8zuW(ixmt0LZPgJWk@d
z{icfBOEb`>Q9o>PgK5pbBmZ<ahFd|WdtZ{h_Qpc%I@C3rL9=x9IkC5}7Q5FZPY!lg
zr2a(6;#CrIn-EOCX<o!SkTEp6Ct7>q(6#sb9!K?)hGR56`~0hO4Uj>(oe~9cXzjOI
z@8DIB&;kg5vU31ee>+?6Wlsl<2SEd>z(u(3+QMSN$2(Ueb({d<LOtdcOe-<CA>N3&
z+rYTk4XQ)Q2>FPgBAET0XNc<~-x|ptOTPHmY){~g;d&M5@#9sJV^xL3w)W`0r1rar
z@ivhA4qA5#XwuQ#>+{w5PNCCyymno_(vpkJcluE8Gv!)DZc0XWIqmIqIKj-8o8a2l
zQK^-JAiL0!Aj({N@-_$O_VpYq=zPKt9(#G{7j&aM`mPG6W5fK9-eIHTWt}f(CXn5B
zl-oXO_2}c-?mr^!$s>N#8!YbGQ1R+Y%cIhsn)F495j^HZVrf{0+EfQb(e`8noj4)X
z44)Oq?=|J7$e5*y3vp&8oomN8ag8U1VEr1#D-n_eWmZ}(U37j*)Q?F(R;Y+;DCdG9
z)dfZQ^{y7=C(1jb29EwNFZr&;3JJjC_UsQ{GwOjM>TH~Pk&&}$D~{`xW|d@HSLNz>
zl1uJRUE*c62`bDR2OzMp884Lux&FSBNn4%Gw^A`@3drV$-~my#si-Dk5&o+|6F8$n
zMp^_Uj&m%FW5(wNIb6F3gyJj^(#NyujmhR?h6YtvJjBD{#kC?1vk;~}A~{HD1!e35
zr)<lTsHZzi*h;Df+Bo8=jiSh8YJsJ>#?1N`T5vhXTFv<FkaDjh+BUh2_dda(AFu7J
z`kTIUWpz|HvN`><M0oMjD3r6nK8$+#_pA28t|NWn8STv6GUQKZy{b4le`7kAY)<%(
zzhdNVD06D5<j9HKV&#6x-T&n#s+r$yT<b`OT_#!j9xCP;7mF|RO3An?4r?LYVx!J~
zcceCkN3BfHCw%p}d)4ViXXyv8jgfC`NX6d!?nigKelq7E!B_DW?1aao)ExF{qsy7_
zHQhE--fFcmtKCLg{P~QtPb_L7?#te=yq%_@jr-l<>8l17;thuds=*pF-zA(=Ch+n=
zAXuJ@P!P(lZ}Q^76+^c3nY2rwO7&%bYEA}C6s+HIDr8HjUgqL;mrG;WM7^F3QNQ0_
zWKaF6=dDS?`o|=dOOE7ZFOZ9a@rHseW++7oEXwVWXGbIAcWfGz)-UM5ydB4c(+ikM
zye0wAvmDp&H14mgr7{;Wwb@ii7$IBl*KZtBs^%;)j#@>BTUOtR=P$mRNnLFvOwH)6
z(FV7_m{3qjOK_Enhi2o4eHQBu&rW)w9VbE)56JOe%0WPPNI>e~f&bJ-Et;2#LP)Ha
zbhNmEXBt+H*PLQc?<LQXEzIC@nL373LEVgB89%7O-&Lk>{Ny1K-eRjpO8ORh6zo&q
zD|ITi&wN*}i5+KaqI@$QLE6dm5Su#b89#*S)gA4Doo|#6S1bS)*IpHn@h~v9>x?Q;
z@fK`LrKoF@6R^7o$)BHDztxdna6=Rd>orX7G=NBkKTfDkN5e0#H8S{gqjL-tz3Yuf
z5bM(e;5I7eL~R{;Srz%mUwr=MCOCEsLsuS6eU(O+9?gBypd)ExtFmjj$fI@c?6WMS
z`kiHx7W05XapSeM-VI7x#&l=09BHTP<~y#`c;FzDKna<;&SSti4L)3~z|Xv}ygv5W
z$*ADf&Uo97J<2v(&u$m6^a4KnilprRberF)axOHc{5zMKCGYWmMg%ihE~MnkLkkO!
z2DdWS!*<zP>C*S(J<IV?!>b-OBwtk;Errj6wW~TC9axFzW_A3aEMH^rOQRM}60_9?
z_(q&abSB0tTpr62i)iT5`Z!#g_2)8o6L2$UYBW&4@wt)Mgqn+&CfO+=FY#$0$}|d_
zrU__X%5mznQWzbZtQO+eV-jV35l7Z<Lz&hQ9YbCmu`n-e=E!d`e5@}v(qaboyUWYK
z@}#55Au%J5KR!MtoxgQ;i2reQRu$5=A&;14%JSd%jb|w<b2x+LDtGY-7k3;QPU*Sz
z*<^`jCKw08e?(^ujEvXz@F#xuN%CZ)k&Sz|?K6&-CkoO|%z<@g>t#MtgVkl>me3OW
zO*FiR5Co}NbXjc<FOk3@_~!C>D|BF`ud<dZ`8=UjD;EwMOXB+@ee1K&U#pG9O3T?!
zivW?C3&-o)DUba)>#C79-4Wkn6lR~DFFdHfdQN0!d5SA=o~I$@9oN#zS8(zj61!5W
z%}Z+AAZteN=U~Ale_mbOi9KC}cg%MEOI#oWe+<TqZz2Oo*|9_=kn`bvi25iAYkb*0
z0jpX7!|xG?#fo>NOAm&#yG38S*Y3z8rok&DxE@!O-7s3JFu>7#*?JpWQb1UX#{6B1
zU)|sN=sMb_$)BY$Xx7so$~yiW&cMf!lL{ON9m@(sj}kO_p_j`zsHct5p$E4NClZiw
zAAtT&V1i7cs1oPjg$rh08C$n?wRmDwIHg0)G4)S_15K{J$j!&)ESk1wE8SaLF<V2@
zQZei5EM*u4P1}s0P?=OYf1~j%VDE_OXKX%RJ{qmbankc0J*pD2z&P3_Y!paXBO9*C
z4IuSD$6c?r6cc2v`$C|Q*(}T0551Q)Yez3qrm8-Km-he`G}67`wuvtXq@fAOraIq<
z*4D-2Pt+bI(U&t6pZrQ=SqWZHOFK2MgsVsOS+XTD^9^&W#wMn%EW8l<T`0_1>EzYc
zH0cs^z^i|XfzLE3VN2Eg;$L2DGed==v!<j72cqcRjKdW8sI7y;@^$R8)J65Lqh=L{
zIXqVP6Z`In=vD@5PV^}IL~-rMCGT<(Jls^x@EwF}&|&>G<H{=;nM(ssboX^4?Qdx?
zo*#kZyCsLuR~RR+{OVRWR2_7qS%tD&wJMxxU^3sFA-MGj_JJ(3)V)|pFR;9bf{cqp
z2l45h0r+ak)tKtO{O%qTz9~3(L}Dbidr1}>QKDA1J&=HAUk8&NFFW3ph@&U?mD1wD
z4Xz$%o9VwO7RdSilFr45vMYxf;*{wSN><oYA6#J_yQ~#Y6};hPgU}Qz124WQVL+@{
zE~5nU@W+y^U+E}MU9AD!1IqzWB~xa+tjAQfuJ%p;v@yr}sEVh~viWIwl}(F#1&fN7
zlEzCDan3}+oYBmC_#U{=+W)oeFw7a+asEiguRh{oYvJX(lkQ-ko`S5AB5xGhR)w)&
zcd)8f%d$BJ-qc#Y@`o`{2MiMqt;QZ{@NeJb5xWk4S*V~=dT@7C`uD|YT9mF<si>u#
zuVzJ~rI;lfJ)g`2tkDRb(_rr6J+-bEi3e+KW*lvc8>R2;csF#dH~hFsec~GU9(o`+
zHM+DL_~YW_G}0>20ghIWC9->^GYE8huuo?bs-z0#^8}iTg;?cb|8%mDBY}ddT$nl0
z7jw>TbhJ}xCM{+dM)vI&h4jR!k@AD4vr8>+w+y-?a#sy3E$y9bS^XWA47^3mQkgr}
zU>XeIrz_cmfzF(kD<7`st#n6Y<1EFoc13$|^R%e@G+!w^oFpYYuG27@_3|4xRo+0G
zN?>g(@Dmr{vzYV-7x#7BDPSd#Ge903Lkk?ps%gvHhfdJCa_UqR^p$w;@*Jh7rzD$4
zb3gkS))RXec<WZ0-}$qM^u*xsJHZR^cpaLHOdYSyz1i>Ita$egNSBv1tuf+v+O@4;
zL_w(ka)~m&w1Os8-&_B+KW>dr@A^y)QgEN&R#CpJ%CKXuf^Ky{5VV-_B)bLXa$(^0
zOT+u8Cb48m)7IuBiC?J270pU+Ed0X57QA^g8Ns!x8Mk3S&8!{2<qW+Q)gYR|@BQH|
z1wV`1F8c%8X}e(Q{rP>ducam!&pS$n#gg*4(altxcaG!eZG(-cO9N&VkNiUUPD#6b
zR*95z+%plZeR<|Jc$IEBssXq)zqG#ckUV@6IrSrh8={N{*Z-7(S7SYy#8vb)BclU+
z)86r|P!;|ZrYB?Du&VYF3VqjA1;HOfX%_S)&RZPzr)2h`ciglbs3$Q&cxIQoH>=8(
zK`{c#xcryT|4&gh8=l2lAg_Cp&lAaql2;JQ!Fp(~>&_r>%5goGqHaIeFc=3hr*L0>
z4`bCiWYLDc46eVh6&m)ctXN3cEXgzgd}`IXkmP!dqn*UjE0Wqn7a~;6ma@ej1G55K
z>@G!pOx?kZt@p7wk&QCve8A?~HXk1EcMPouAw`8RmagG|#Ca{aYeP2{RMIpzsX%K>
z;zXga&?;EY&^;6S0N3!%u7aP|ZPR+uD}}}$C1_rIRUB>B@DN%(7b34?qLng~TOt)x
ziXcjge15^fZM|c6`0S%oESu1*DE43hM_0L}fE<~qZGFb}d`l&`_oe$euOGZ<IOJkW
zG&?UkEBd<2gTuR;i{ohTUQ8+_#HARQtagRgp5%n+q}{n^&AJ55eyRK_4}XcR&h;xS
z@1B?LeCUAks3POX!u_4h>SzbrqYaV7!%8tkXjt$gqks7`p_$-0i7%~HoQ99A@HWe7
zLe!;lT`q~&<B7_*H8*>z-p08Qg4cC4dQyRuJf!C%)rLr}sdAPXRKN7N!vSqIQ}4B@
zfDZ33&V`FTMvG%N+!;b-L|wM~&9BK(?b6hk==dk9Ak)XCugk>Fd23{Vw4f<~qsjy#
zaWT(kz_IM==E&yAW~ic9PcvIGU1(S8`4+I4tXFc~1<U0o))~x+oqBpu<+ka(X<kli
zOwXdRpK<mzKXzf*jqt?k>w}4e{`w~yrHX|1yjMOwzK^O6wX!LX6dx|~0M7512Y}fS
zK5otxE(aStrdk_1cP&9{P$(=f0<|#mRh%#mN>+ILJFhFX)!BCHKAHJZ(kNvKm6kcW
z57}GKiZ-67ryGkQdUo+I3*FY2ew~Q(KC<813!^pJxo7xC$kV2Cyu3FZwMTgw;jRuf
zBl6K54lE5~DL<5L@vM%iW0!RfsYTx15_D_FgpbiMRJ-n?4=?dB4pvo;X0i($ztD5~
zW-4fwDi<jzhTxa!x|At8e@6K)>_+j5l<x=8v!maQ5nRjCO`eAq>j^s005v-mycHQ0
zD_(NB(#)@1E0)$hzi8r?K~$%?%)HOiDKzLEfpD5TkBI&(2GAvZaUO}~0e}7jd@1;v
zj@&%|16cj*JDF3&FGHR^y1&?Zn=luM=llz~!6T9L;HLUN;IKW0L0Jfh`WD;Tt0uiw
ze1D;5e<?ny?oh2I@Qh*tZv@wh7NNg_aqHmnx8bMuLlv-}wRpZfEdTB}L85M)3aumP
zQGpkQp3~iA*W!A|!{}3UlMqe?{O~J|w?rUA-l1$PJ=gGx7LC;54b08qkk%oaXU_uu
z-B6MUzXmm0hYBha7!vi`8|~T>+UwCT7|Jk6f#K14v|tg4H+cNW;j_lVxZ6Cr#bnsH
zsb(Dc+na3}ibgp|*SORD<`;1W7=VM8^<M`~j9;A+O){K6JD8>6`)WikC%A$VPz4pM
z`P!vy@f!`S9<35__(po#@-fCEwNBcF9L*=0;QO<Sxw}I&^=;ZQ4ZnepeggoB|K%!>
z*TT97Yfh8rTgHRI(FN7-ALVvW>bQi)uqKoSp5jeLj%UXo<dqt8$-Q_DB9zSv-;$D9
z{QOw?HC;|$w%lY?6kBLJ_JK(cD6LDh=XJCnyR0&so!pTjbGnT`lR*qvkHh*)QBSF$
z-ZV!Xu0TTVjf952JYBLF_na`wsjRqAW4KAy>sg#i(P&2$6+StY53x?vKrss}o5&=T
zd$>t2%kXV<B|iJ=(CSKKFJzxXiNW{KN^=8r8p+T?lW3xI^VXI$oD8SBV%eOSTTw=n
zCs+mw4z}=??@cOE>-}E#4odqJI{Z4&=-^G)d}^^|l&j9lR*bgPIi#8Lzw<&wBdx{=
z407RVH?Urtty^DwvhKOUo4Q$?1FRoIURbZ2?RF`4b<u%vlL6q-BjViyT<$h_EqSs}
zHc4@8c50KYb~3+X(wUj1c65<RojwE_P|jJbz^%thvLduwElz;{3e4oN+&YYnMN^)z
z(0sTa90=BA)lS{$VW<6^;;9;-KH5EJ&*W?)UWnyCqoFdsa#qQJaL-&`a3J^k#8ML^
zFW3cFiF_vvLV|7>L{|jiAl4x+ci-WcO`N%(#b7_5HVsy4^klFKX4UgpV(dQS{#+#q
zSJv3o_c^Mn_+cL!Ii54;GpN0_1ZU7-kO?2}Tialwxkk3<JQ$*;3uwF?o?s)b=ijZD
z2&Gf$Qda4gXU=aYJHR1?WAFAHYmC5CM+)v61PFN#R~@f(G#;JKef8@P7C;SZ>nMP4
z@DW~#6_jx&3J<Np9Z;v!?Wj&7p6mYUJRK5kxENEok$jeFB$guVjfNjdceSNwGsn3|
z!W3MRaKxiTrCF7+>%inm>c4zM)5x5t-XEc<O|iees>tV*55yF2(-6u$yzXURJ0_%e
zj4(Fe=Nzk#mKV}Xkm_?Zz>-NYWfTjq5|Uy~EpI^ane<IOpDPqKOJ)f6xsp2Hm#NyV
zw3=ZRx_n)`$yIy8$akz-`g+G$fk;Fg9L4NG*>E1^6r{ihOspWA+6Z}}aCrg$;_8?j
ztg!>kE6^}Rk0r2fp3vYWQ9p0uRAri4cNGtS093X8iuxIETl3xuiQuDpp!+iSP=2gB
zd83ihS^m3-E203uGq^r?22c0yP4wkaA#V21xdds0RE?`WIqk;5m}Y!MONZ@PW!%}!
z9+sSVYI9Ly&P!zs)zn<&CuTHK;m{Lf*2NsD_;j`Jesa4NDm2Km8`YQ%(8SUbeC_jL
zEVSl3fMreSGDwEIB1@-YI0xdwoyEsAoI~sJ1qbUTZJR`x2*40+PJI4F>)?Pk^I*@T
z0mTTn3H1tEX=#W)Xu-68!(nnw9I3LX%Ge*Hrvi8A*HTphtDU1JM>nyx6JKviL`Gei
z(KC0A%}+Tj*>P?(fFPlj%-(yos($Qtp<oZEGQ8t6QElo#9bK2;3a(p?{a>uz@;-iU
zfClE3ip2BkxvuAITejL<frcGJg#$qqc4iH28ZqUcFc!_tU288ez!`Nm={1@3u9W!M
znxTtl4K=G=yA2W`Gt1!Jh5`n~x?QkISQIQi4=vZOsov{}OsaJ0()uqCv3J$9WIh>c
zrBUIimWH3vH<9E%UkSuy<fc67Rvc5@cNI@`wYfC3Ht~q~UQDVe8@3PK<D~JgTYnr~
zfidHrX%*^6$gJ8OoJx;+dJeJ*8QE=A>o&j_C4nl!$B$b5oaPXtCFNUayX)Mk!w{rY
zmBMVs7DV~OW{y6(J0hXkV|3rFI|ypJEK*IYJ6+72|Dm*XzNx~S*>WJGbc%pvBKjF+
zfBSZH$(GyEqi1*ePg+HlV`>SZ^T~^9XTSR>sAF~C8Wfy<U<QBP_X6u>EU!;X*NsG$
zi{CJ*g$HLKDF~xKR-RRsj`v?t$obA@;XkQ3a9&s+j#BTCmPtld#PqPHm!gIbAQki>
z<8;Gk-VQc-I>8&%>(uTtmyZ2`yV*s&Wjl4<yg$Q7jy1ITKo*w)%7Ps~iz~R^AI!MH
zNH<A}zsVGM94(Nj*5zyswcf1|lrGdf1L}SJkCkww$EWev@VpARau~Rrt>R@)Wn!4g
z4=1M;P?E3TrVSxs4gQ8QPd*@qD_z~6+kl#ns5+AD9GLbFKA(e#tE5%+(YrQgU5%>-
zQ1h&`(b0pS<1><6AKZI;w`)Affl#VyB4zAireP|C*-qvF&wYiVc9-Q_#VB|b?zj3z
zK~T2AJNF|3yk>CZO~!X>wf@cZCip;eENX3><P~<uDVFDDCn{oHPJ<DfXRJOsiVlP!
z>5U2dE;;gGgko+4xf{fpoQo-ES9LnMnL(pa*bXES&Gl52<U4AQiA&9j$)6(kx+xy7
zBGWeKBtn4d>Vo0b#syVv_RpsD>yKWU^vA@c5?jb0T)M{Oa`^lFcw-l!BA?E!|J6ma
zQi`k8uy|`efrnWE5oeO+U+#RP#T01C=@nE}!*82ju^x5S!(A4cbnuJ>L28ZQn{7v$
zE;sXv3z3!xh{ntC=k=Yn#7+5Lo~Xbe6YWEiwlX87QGI06Pue@XGB@N-$2p2fuy$E4
z^K4>B;z^Pwloj60X2SKeypxxO?UI>2@YL-m+lF6~{@=HC^p!TB!hD1r-@qbAo=>nq
zZ{E_SII#7RP%;>~4{>(W<1i`C@GvL(T#dID3YREea4_#D$8n-xG%&36DwT1L<E<g$
zye7513TeDG1KDyw7!$sieMxL)JTQ%eqR>oyM|uzFRh`Y-N^k}Yz(O(v0FXR=7Tz9=
zHJe!!Uj{7}_NmOOKc8CVkvQFBP#hLxe(PBK{U?50aD!c>!4CF2kj>_z)oznaJo_t!
zLR(;sl)x!I(6SGrX5NqHu1E;mpcJLwAW->z9b;u_xYiZ8Tm^s9pL@0+b)acYgj07!
z6%u?zh6ft;k{(GMfy@o0Pupg=+m2*4s{ZJVL#}RC=4f(ASPDl}mL8Bkr8_95p-)_L
z-@vGNt=*Huq?k)kp$z8WqTuu=aXGhQCJE$AX;~-p9&~V6?vH-wnKj<ptzYOTQZIh}
z#=fw?c_~><XG?Au=DI<SuxIzukIOmA9eb-Czd-OWgs8&1!bQo=zkrXl{J}%{Q~|LY
z(lI@iv^v}D-$OY`?H_Ozlb4g@%)7*xOLsyO&`QsZ2RO_;;-Rt!I;SQwYR_#CXMb|J
zeus8EZ}NsAyj{Bd7fnxR!6B=uDytk_E<MNEy!EV>n1zvB5*mz4lx!88>L9C&d{sXW
z<P19MLX>zWxba+yBpR}dFO&EhqV0|dP#Mm>FTZe>ozA_l#%$keAc!tQnrpITehph*
z4tKykw#i(9Y@!<|-8^GYXuQ^RC^%&LkyR>WCm>WH8m|Im)!DFXjo62o%Ni7Ap^jo*
za#LR<4;upFyN#c?%V1WORlgwzJ|$b}9WEmp;dgh`HvFCH@QKIoos>)A@tghibal5O
zdhlKk%kZewbmk$RZ;}n;8`*zv)EuzFQb_do&#u|#e_jffF~O2nKuw3H((g7g69g1*
zma{bsSQuOpBB247Wm=095(({gO<cS(AkRw(Lm;k>kDQiDj1|_3v^%Fx2zMfbeNxtp
zV;$EOo7q=rppgXy@l@~w*EAEAd=pVQUwt3wDt3C9^mFQ3#YpJrwOPoH0%WMHn|VT-
zJfg<WYsC(8H-lji0Uumc(x%$2o<gViodT85i5)sRj7#XM@6qShKR8qVJ>D%f^jyt#
zvGs8jTRpp8@sXwFjLbf;=S5md&DpZtvcKxc_U#S|l{y~7-V)dG{M^y}oI{30!EZ{%
z&vmV9YS_I<v<NF)2%6xSa=8n=?5@sx++aRtHCH=fKoh)`9V*$KD&(EcFnGCTi(C1w
z<XwKS++|LMXKmEMm#~2sDx-OqpSj4#7WfS=>M1#j%1LQwIg8WzxZEcC93G9fw5n~K
z7n0_5?!@q`s4oa~r_jQ<_pmQASOk=Gp0ZFpDV7*fShg#q2~^N6z)ZHkU8?<O-zM69
zxFCK3-=H4m?6cV<-epwLYQtQSazTBn*!xDBXtt|Rn0w^&FKsTA`FVEE+md(k`_z@K
zbUPEY+?L1qCSh?g2E^u*v)&QUB`qhfaOb>~+Oc?6kPoIpbXyYO(*CSXUsj)Hz4aMZ
z>rS?+v>Dmx+Ml6TergtU73z>9-QA~;o`5ju$!s5taPsHl{l$#Ot_rHxg0u%1sS7c$
zz>pOzH~g&Taz`p&eb(L5O1NVSnXU$DF;db*Q@;v@U+c91g8*S)%hWJ2t4U{hr(pA<
z?)0fa?3Gdf9pYi%2UVF2sKMdj=j%&(j*hUqyLF<LSCIxI?Q_HpOhlv3{M^xUc2t|{
zQN#Vrf+PLD6RG1HT}pKUISSjvD&w!DA?NmTzDIbR-q%t68tTY<am3j)jc4D@u?eT^
zvPtW7Vs#GA2BaJ4ZEKCs4#>b5^6R5vP`Mo5Yp|5ZWM)1}YGH$XpQ`fiZqvv_!2<5T
z^hawGc%@=5q)T8ug2d#9zsGrMy-b5oda7XVba7mMZ$F>k%|!|{BF?I1YGf?LYQ`D2
zkKzO4uNOXCaU}6P)96U0^2mITC+>$}m`C5f<vke+rb+MfW=sA7<>p7$dZ^YLYYy?6
zWs!>;l@N3B-KJOA(RSUI3d&b@+Hd1i3Ye#UR(jP|S?z?vD{j1!=@@LjtCtN}P*vb@
zwwJwU$P}I-+#QW^zmZ*6I!c3YqPlTMm*_6$X|@SC<snFj3^7m~02kcc!jK8LX;)|Q
z8onvH#x(X;8rG4&2l;&Hbb5n^*-+P+T&fv1n>}y_=dh|yA=8s@>C85o#X`0;8I^d)
zzr7MNJg*p;mBv>N&)W;Lg$1iqo8F^t7^@sdI5RYhBs`1a<YKJqcP%pan{<h}so}Ge
z-C|Ekw;PB0ed8z&8NYIihx_|w$vRi(+$46ZZ-CKfA=GV{IrRWupGf7Fl)DNWO)u?y
zbGm!3(EmfK>&=ou?LM(dBui%Mri#IAX6o9I2M6S);%PR6nY?COb~nQB-{$o;o<`ZN
z?cgy-H8a8icCv>4<bi9%IciIYc~fQhQMXbV7PydMFqwO3rZMMvSSg2$>KFcQAk6%E
zb;XnZ^d1&G@df9apQA%j+sJ`6s_f9K6LvgqF`bF;FN9Q;q2ZZT;bNGh!~Oh;{faW)
zS+wFPQ6^d$^5WkJZqWIR#%f<Oq}KR@f8s2t=cj!+TOC<PuMaRnWgR1auEBeIHz|(L
z&gq#)7llN68F`O((Fr?s>aQPcLw^KXDE7Y8+V5~r*c&y`AtB)L&uq(`+Uw*WP|An`
z9zw~F5Fn37*v!Pc*06dWBaz7~7f83wT?)@p?@O5~dS=I*3QP^PhV*Zi{B@C4EM_Eo
zBzn2MntjZf6f@-mX>%mX8}zV>Y@_vW+EZc6N$(@3YLvMv*{n&FC9;e?)_oGt))_49
z_z%^K)lBEc;ms)I_EVIoOZ?#;@8R;jGQPMj{^AWBxZ`fBaR=``jY_q9N-yP++3f;`
z;z)JIT+RZ^e?f>#wSyt{9|^U-ab(h6Yp15<%bm{9s_*TCIEt8=iR614Ani^u*_Afw
z6Xd!6*Pfdfl1{c3H<SXpqHI2A88IqNPsbWnZnmbrC`(GE^oOteRK8l7c4heZqbh^J
zJ_!2#ZGr?mH_{;IVa)TC+SCLusY5Dyewm&$^~TAEN6TI^E4w)*vl){p#S=~DdmTNf
zXlLHl<9(Wfw^h+4d$j|Di@q_u3GO$-&M9e^HWM<5hpb+EpRXV&GKYZIBp|XTe5@rk
z;wDLH(QHywOxbg4{$WX0X{8N<r3c`3QiH~Pxs1MT9QL1a<Zr4?B(C2Z)EBFzEXAUR
zuqj47MVIXCE@dwhy^AFkW))6n7(I^GnK%xxd2ud&t$d%1PH0r9+>LS&gv3i!##vd=
zplsxcGGAe93`AuF*2dEAf@$74adWySQt?Y06%bVeX+`o@kVglZPP^7q>JqMLPNKZJ
zv>IUp(1$TFyZVb4m1kcFjySiK{*RSG6}3s^<Bm>ZIdBUD7>5QkMv6!TBpJ_U9FZkv
z0v`U*Wu<HUSU-VJwx<Xh&8Er(<t~*4GMFf_zb2CDsn152UH<w!z?{bJ0n1EvOiQ&m
zYK*tKe}f8@K=MR^JOlRq?dPb5!o2KtNHrOcpN|?-qvTz9Ub|%^xJAyoDMsz+sO)p-
zqLqabX?@}_kN4#AJn@vI0R0WbP0#2-L)1Eu74X7-R3+@=YhOR6)h2u!oNLYdMH1+z
z^s<Y}CLGTJ$guOTy%Hz%5G#cq2}+z;V$bp8>)-^Rm)rr#M?Own=Nt_rCg&!iUv%=g
zjc?-{{(>}t$UUPerSV7iGnAJ)BG(E)%jc=lF~K+qO$x!+FbK}a|JgeP;QOiyrx$2S
z`TI!t(H=r|iPrr<-(upE;iY1Z0lC<V{{iLB@ba*j$vz9KTD)FY2t{T8_MzV~IQPjh
zcp&EXKaT;iY!}?!q+!A@ap(kDl=P#~0kqKgAGA=<F2ac)2LMsObaZu>f4sFZ5B?7N
z!nqA~1P-<TWAB0RMBs+ie{Zz)Cl7Z4%?9CM!gVfl(C#ED0#MoiW5{J1hO)AV$7xhY
z=JjJ?f8W1y6?|gs+(8^2jf_?@ZU4Ue3{CPj2=ew)eXr^904oE;nyq8&|HwD<pAgOh
z+Cc1)P}X5^?obU<<co0Ie;Vou!FK}kX%0(2$Dvfb4Vt{Y(z+KNlumb?R^z8G%_v3E
z&;aspFJU#)9s>6TbiNh5zRxd3=B55~KLCr4H@>)5rdJaFG6;vBgdT?{eR*Z5*`LFe
zB_oQ8@;%FQU?{0dJ$YAFsc?eQS_7~u?3eb#jyLn2E%}x(qtWt#@9-A({rNcCTBmlM
zwswt5jHa!PaD*lf;1xz&b{z29hqjeC3KY-oyG{D9Z(oZ1yo2{BNs)j#O7g`tyrgQV
zjOO|bK&v0gcphSLT6__H$%secEk)s}c)ip`C#P`#yrYXQ^bx{^M2x=I7tI$p1R?(r
zTOIkkiNaU-KGVr|W`oDGM`Hel11Ud(lKIyc0wX`XCoOn1RxFpu-SW8^_x7VOUkE#I
zc7yulRqL<08ZKGTeDf4XOB4*egKxW=^HC7G`joe2KKQ9$aGLSmjs4B}(X_3q?rU5N
zzaK!319^I8o$kK6++>&L#iJ3RH}R!j_S-*x{EI=|*fx;RZOzXsxmxUuj2ZVHzPZ8e
zzh#NZoZwWMKbN)0OPsqIX`j?#z(4PUkDO%W#8nH=R^ooLH|MAKV4n)S6#9sLD+(g`
zjZ$@pNB61d7aS-H)RvbAV>KS6JbQX@@9I{Df$I~ad@`TYM!xR~9j_McmrqxB?g3(a
zdnv{e@hV&8zCO3-tF*e~_FqY9^m5v+Y}A?8^qvK>uHsS(rIG$eUWN-|xt=sbkil24
zUUg3Uz7$H+{pUA}GK<)Hy``+^Wlq1Z6pA1JbN^se<!IIej{^Jr?*9m|IxHg$vlE-=
z!9UMoEZTLF_}<}h2XbTlf&@La&bddz#A;R7Sz2<?>^B$qhv-j$n!#9yfv;d37v0^V
zZ#C2`#PsF9O{iLgH6_!|szKf!zfU+Z&u%Y5g8M<*R^Mp|u>O23KEMzEnMWblF$7*X
zG2e9pSL>A_Zx5vtZklAmyH1C*h`p7fX(upb!s_DN^=-rRNc>)8vSum~UbfLRYI1(x
zx~y<~eB08?3STsT?uQ|HAMmUrGFx=O+dSPn@^$K8txxwO^kRv#`56VOMi#7NU+Apw
zL1n^46Yq75zv|u)+oY4G#*@!Y@>v_C-5y1|ZR?R*n(#tjfO(BCV_|vFU|5HFZ!F&7
zTW0XS0d*#p$&C`^8zLNfnm5`>B3u}ts$UeI;f%}3$<><??ifkhQ#*V%O~3PGsyvFo
zOH7G-Fhcf!+m%~TO|YXgsfezdm(UmUlg~&u8~#y_vcwdEJ1o5;w)b9XN9R~<?tUW3
z5lh;hIMW@xfE<}uG5@W#2rf6sQi>Zsar)o(!i$A<qRht~4>~3FeWq%c=`)O<;SUs!
zY6``U%N32%zj$95pl-2z1!8hnvv|F2Or}#P)#@oi&6*Rh&;@>T$_Z#r;3~MUn6@XO
zktUi5{>9H0IhmswFI2m@;G*hMZ52zNz5GtM#y%+NArmL*!w8yj1!xVCvQ9U(@Zlz4
zGk{Bx<`cmFS~Rj*nJ07j=u@O6%xLpLR^P>!Y|_K#Ohg{3oB!Pp8qm)JjQclF{*w3G
zGL8#ac=Y$(PE%JVxIlR!`{%ZZC)CB~aFuHho!+FwQ<b_)1I65x4<K@2KTY^$^Sh^$
zdQ&tMUoii)gopNSjA02SsHcgaDPPWFiK?kpNmO9HG#rj8R@hqexDlN@sa$^l60jFk
zy-R6W7<djq&Jz-P#YcD0fq4w5MdF7*_qWocn1&+^L>sS9hqz7_C~kEou9A8jHhxi0
z<RYX<Ifi;H(@c-kY8~9Zg!U}O-o;OoUa=FXVrrWvO!=u;c)#$(Tv`oX_Ro}v`NeA>
z;wwT(ujmG(|GxWYHvKu|)cv2{Z)L53LNQJ80S4Zzu&>CSKO9oL1~|#~_ow|)@SipS
z0%QIr@HW;PFf>AG!s$S9&U)bPoj;6B1NKufO#bHOUrJ^Sn_iI~==_#EKve^;<pCOA
z1SrM%|BWQl-@+uOBPp@VWkaj?r+5NoTrEO{O>Xxf|4X;*BJtqwO;J9olJp@B#pN-M
z6Fy=`n()Rbs*=jp_oS!4p#{m582_U1m+{Uf#~KRfk>&SYE}%pm(Aqzg0=b{We*R57
zP#xfwc-mUDm5JPYK2*&DIf(mz`fk4epP#JHBFu59W{g(1|8((E_AlOjKtmU~jX(yP
z{sJ7_xn;0yn(&`3dO-F5xQl@YjKXK&sL99M82-O%LErM_jV~JIZR8kms~I@8`=@_e
zbmFi_r+*Rk@!p?-KnG0GpWrD#?O6T4^;qzTiJNQ-{^P9Y7t`EM?6t(eRy(zNxXesF
zGEGH?6wMBc1WT{WdC8p3a4ZMGW&LYEZ8FajI0kU_&c)ekcBjP!4&4=E=YsTfcsW)t
z3z-q@@XZe@UrT8<z;Ax)_K!Om-;6f^@jK>W(o<JTbNKZ&O&L!_H_}j&Q5YxVo2-3j
zh&)dt%K@Fn^7FONwGAschZ#wW_plzU?4x+Awx76LvKu@d=Yo6rm{>ji)D6_R{V#JA
z)jo)9jCV|#=^n_<Xki%+<WV@fUw<!e3Trq^tVY%v)8cQH^25;HP<bCh=TCc>F@qRt
ziXl`Ja<JLvB)z9`k~$};`PUu^+GRBT*Ni6z?fcerB?zsmVp4oPv!bRl>34CuB|iPD
z0w&tJPU8_IK2nQTr*LWJ8E;Jx7IAV842f9o^Wu65cOXr3_?}X8onZR7-oVrIc-$0K
zvjo&g$<EK^Lz!(?wZKs^y(lF8lv05yf8&vkX8rDu0cP84rKc<%?aA*aN$=3%{5_(A
z=oS(j_n++m69%Gg!*^y^cQhhskVXE?$Zk>6%Q38hqyMV!N9=@W+>avjCa9`b{Um$h
zWsy4~Vt21o00v#7x$M9XU&_t06=E@Z|C)r*hpfZId(>5&7lXrS-NnYZU;S``X49`k
zm@KEb?#Gbf><rzt1LMcR{NWF9q8BPNmi+<W9w%bQSTwv9yidpxlMF-9hNA{Rr{peA
zL#JC&3vWWY?LW4)-#flQmVG4HJ@ASMU|~_>hNIkUpMJsx4j}{D;BO&J`-&`CYwjWL
zI4qqnhWNdQATplR7@Wz#O^}RnrRBt}y?Tu*6p!2sHBM>_+EP?)KdxSQS;(w>tu>w7
z7NSW#T5q;Y>=MS&*t%E^F61H`B|w=up<mEgR8DcPoVl-c7bems>DoTqwOx94-eKbQ
ziUXe>d^~G2Bp&BYGGV29vroK>fv#&aK$#eN`7v;dCDOI?$NrvYWoUcHXRpY(1FHU?
zC?%ASC#1}1brQ?)d=&BFjDgZE@!tFwV0r@@>(_mZFTlP%{qnhk9k~t`MQCbAu|9lz
z&sTZkMtS1I>tv#qRYg+oCPAzEGR^2J3YK~@P_t%vexBqPIZq-`K4CNFj=Y^?Jl5A9
zC>CTZEwwsx=VV**JLhN7RW5TH-7VPl<{$f6IFW%~>ULfDo&<~V$Cj+hy}p*0VjY+S
z<tl<Y96UK<C~(v34CN}%VTF{2b-wZ+N&OAQ!n$gew%xxx5EfU&&T&FJbs}CDYjxkI
zR7CSFPpjgqPm*VQCP~7vX~X4ahpklKBW8#3gd}15^$w>W18vVb^MpekTq{${b`U;L
zpBJT{v5Y6)8yt5lzqI*HV5DB?4^OUV$vykDZ}StF*bd<OJhwPzHt-WTK9l9hxARg+
z9f)jaCd$vpJh5k8u(0#27n{jufR{v*Rg%tT2wT@?|Gl8_#@m$-ykR<(fjvrb(6%dl
zejNqPOm#o*QSa`H`Y1nnQJOkgR!>PSK3*UP*7xT&0>!W3jNCDOx;#vVt)CZ9FFQqL
z44&%q(t<%h$;a|q`z$Dv!d<^Ai0A4|C#`6`Tp1vpl@o+iTkl#YUrf#QR5P5Axp0wY
z2$-5aLjZ!@^kR)E38urU>(9<X_6=2apgAP@j2DpoiNi42-$(CL@y$Hy{8+DQVh;1@
zJhcYU!h_~UzkpP@PB*K^f*nW-z>XK}AWt^O)e9L!BXNCHlN2)$`KR=kt@>LhV?ewU
zt4|hfvs7*rrz_S$aEb%{inGBZ9eVC(_6Di^jWa6C^5@dW%X#)mC+AcISw$i>WD(nI
zH*?AZw$*37tcE<_N_^tDBg_OoL@I|SL^mu|rA9%9oco^JUcWtWmWYa8`YaJz6RG*F
zgED2i>_~M*y+InRgGy0z62~Q~DmZkl+K)T_^aEdajKi1v8r+oRV5$h5djqDa_VXcV
zg9`$CQh9W2*EYI4+x{Jb;jqkhY2b`1pl%dB5u4%MIu*NfrO*C}9Ul^a#|mbg$5Y!T
z@YnXA61l>4o8^A-j^)fhtq^3vQ}msZ({~-gmPWci5h__&)X{!bzOi<V32k^n%(5zu
zx};v}&UtLcq~Kz2Vsj<9ehI{10VgXuq~=IL`LhV5dhHT2hJyM@HQGk?eL|h337RB1
zft-s*r|-2k1L<tLhzqa9T0qb8?J~gv>RD{RMY^dTs3VHbf<T<hZ*w^W6-GM-s-x;H
zl^Ar6>#(AqoH*d_$LFhmI$epfT@>7oD@;|zoPaWiT4ob8r`tp*NL41S$4{47(@Gk6
z5lE3^UC0L<j~exvjjkUXHNJ@H;j`};aPW3m8$h%7E-vs{z=c|k2>e+1TF;c&x*4d0
zn3F_F4KPj(5RBR!Xcr$!1K|g5I!TfO(>0`(ve(OV6&)wDO-$AAvp7t?P0j0>7IC}Z
z^f5<`X3yYh_{7<%-#ZuLeu^X|fUooPPv8Cdwgvgay<gN<1W@<?ksamnEfb*yqhIQd
zT1#%wc=u~ZEjNPIovFsTUkFzI)75ts#(lFlp&>P*`Q$u1Eaha@2+PR6X=mHbQ9wC3
z$(6U7W*$r{Xq|qT=GcMSKlHM*>bhxhc$`#e)e$~Ncc7iL4XDpPNx}+J_(i~ObTxZ-
zL;E=Hc#IGMhvWJQNz52T6i}|Z4abAYLLh(E3qBUw(?`)%PVtQ?b?Hg*WI0)A7)P?m
zUTzQ5!e4#Ojp^DHyK#I#Dc_7$<4MI+_}Cfe#v%Lh!&Q4KWK)TAL4x<CLRYSLRZv}<
zPv2|RzV{k0Z8D~;hexemD_N*iBv+^-LYmO7`E4_!tQWkR!2X4qMD@YTV;-XtZcPIi
zEMF!TnYl>I$~>H&xn3(=z6&U9O@3~(NAL|@PI9#{^4hM6SI`H}Cm5~a@CI_MFF50n
zg`ct!(-rU&l9i+--V|Zr*ypO;#gzDJZ#68Fk53ovbPJ+0!KrX%zqLVL9{L;SjAI@T
zO4P8wd7|In{1bovU~Y$*eW~Hdj6fr?9I>xbwoy?;q2Sc{!E7FbgP0NDO{uFz;0C%h
zzP6<CI<G!=>jgHw!m%f(m}hJcFk|lbzu`0@0ioM&ri_Drf0(cl;P{~QmOHIr#3~Fh
zdFRW<jvwwW;YEm)*m1qc!Y>wNYi-X<!BlYSRm48+{L;52lyo2iFDhooIA7Y+ygVA&
zWL<3biO_C1{mM@1qtJR4e({rmLbpFt@Tw&hjZ?|h?Sw*5qO4|agipa>fC8uZIfa2>
zruzUQ&6VaDd!A;lm7BbDV^>T3v_fXZF-)a8xL`ysENuf=4&2(b)?DGf-~hmB1$^&F
z#QHj?>%CL>s6j`~gfo;)#!fu?jN9Ef(x6Oj>;(a3V~YK$l7L<CxySE_>W&#h*}G<|
zcstK*rp@MDe-kCbFQEj(>7KC5vgYmKmJ$55q6ei$QB`(aJ1@%=^~O{$99^%YqDsOt
zV#E)X&}WENlD6c5iG4i1P}qMdb1~i|24gWu)!e&aGSXs?+o4Am^XmD^7qVvRSlFFg
zEE;XIYd%*M7{p&|2}&B%dAfgd&}bJ67(%;=RE>`)X@fe2VQhZD+fk8QU8RT`w~F-|
z#SdzGo&O;+!CLH!c1PB65y2c+G)=2#=Y#HEU_^WBRALGJUTa!XtF+VTNa<Psn^Mhw
z7(dA51I$5zdiT;~PNs|0liqd9bxUx~v!R1cP2?uN?xRosQKy!}q(zvvBa)RN$#)`L
zcDC7af0Z`JX6b5%%%F70=`{K2gcc$+a5{%oza}v@(vnjQvLj^pOM>NnUk=mYbVJ#B
z^)9RP{#%nJ!+S`tkJBHe*Xq(V@h2$S@vMF;b@MkvG{t#ZfGL|l^;P#rT-$e&BxghL
zg13GscZa{@lOO!zwALR>OX+cTX;WLuE;q<hhWuZ%XNR0>;nC-RC~@bg*)tKptck7O
zgL_E#)$HXh!|KPAwvxB9>s|f%WmMF6SD+uyHFELKJwNOe-+OvY-4ZCv9KW6HDrRjG
ziWV;W+X>J>V?GX{*ba$yj{Ez{fo58-;afxU4tl<tevruOpC9cLRdX9qa8(-`PLlt6
zt%@(d0QjKV-5n`=x&wHy<)q!{Y&!9k&(2SK3Q*<CfI9OHDlfh-fHD=*Qu3$c>l}G+
zd(!Kj?E$rTJ>#sSMa53AhS9oGGUKGUjeaopjXx*JZ37wF9d8i9pAWd)D{?;s?Ioj4
z{BX74s$Pp1YXj8iV%q@+PYVPPWrO&g&+C<DGq<oFj^^^e(DUV!*Mm>*9rY_1X%rrY
zmExAM=LOd*?W{TGCSC5d5ZEpztYI7t8)%U<sC*$St5aQ>&zV12zzaoV?wxECaJY#k
zIq?^YEAUf|tN>~FBa?lgg?{9<XwJ!{`f3qiqUq59z>FEYym6+OMQObA#nsCAofE^s
z<e8-zTr)Mpi7asFFUq*K;+%=v61qzg#guOosz7oxkbCX$@HOLX*+hS#-I~H6@WD4f
z*DVYS_bo>qbAZ*7O+zGA1zG+WV`cwb&**hSM(|8ayPBc1@yAU8*USsI7rPO~zl!vq
z9g2epR(2NK;q+e<kTCtJ(a!=%ymkB^^f+;8?&#e4-W-$L4gNhwAl;kgBWnXJeF(}H
z?D?ab%}R2^+BOvn2VHsMfMW0C>tH#F<y3^Qd$?8k1*v_J9qUbvDrbv5gU;B)Y$tTa
z-=0zen1NO-$Er28;Lq;FckB#-6+N)}|9b_iDn8G`cp6aiM6Dvz$?Ym<;NkM@3{`NQ
z0o|C+nJhB9*7S^N@Sb&rJ!OBd0ElyRA9rWau_^u#V7~nBN;{p;&S?Zg{v_J-_<Hf}
zh2)W!Cp9eEyQuSG(3qr?Tjqry1qxlyH;la*D}<qoiMJDBX@Ekc3A4x9z=>wRAAy@l
zD+#9|7w${4_b4QacnW=;s!<^<#>_+bK_lU0<8?YAX!d&q=L(@{`>7gCJ`ODfkui3i
zLrUrkrms~~J^+t}o0JSkr+ll?8;jD0LGvK5bb#dEbIpcO7UvIpM;D<i<`v(WeyZbU
z_7<g?u(hlUg){Rgyd{b#Zw&g#RDFoVb7=+^s}~K9K>fIk9c3R`{oDu|2X>n|vyy3k
zoxYabuZ>L>COj9tgsv7sCdn3Zz?x|dYWQ~6NIlIZdZK{;Knq2By~74th#sy&oJ7|f
z1mUj?+b*6N9k>NzRxSfIYKx!m(xSS|l1nGf+#^ja_%<^x-ozFQx7B&lMJ9BPy`e?3
z%lOwSA8#rPhM2JHl!1mk@H;qq_-m5BdKz9~ha&5y<|gC{>6uNFd#)IPmap(oOQeBa
zQsbpM;q`)d4i(0c{ajKxUgevU8(!6(#jk;$H;w}ez#*MVkDAEB2d;crJDEWGg_D5m
zc7F<pTJ9$1b}0qEp6`t+`NLuSPBsP?bo{ki(_rgZNtv*>RNm!$tB$1jYoGdJOTFh<
zl`K=X>tjp;*d6zL0M8Iy^wH+tlX^eLGFNYYPqlC;==Y!-zMw49o9I%i?&!$*QuWC&
zzPbCVix)<98D3DNU}2_-PaMIH5y?l;zQ)TZ9Yw-3ok)fVYt@%e23)^U689greE%Gq
z_|yhrY38HecLme}*GTrkjJ$r+?&O>j1ymi0dh|0GNJ0ER0D`&7hMmv|Z}jC_5fR`Y
ztwoZ)5hb)$u0xn{2aVaP9%`f!)>=4L%1O^vU8c#V;zZXadVR7+PA@Qx3C0%-!*9M9
zHT4G8g>m8~f1v4n+uLIL#RIzapmzIy7^pG0N3CL^38-^sFH!wz4I@Uys=G_T;kdD=
zBoL-d5vQ@>qR8LPUT!=prIE((Ew5Gh?xe+P(o)4NYu-zVE)FP{IUU4DdH-~@2Bs=X
zf~S)4G{`xWWeOI?PxgOt^;H3JE=#vK1ShzAa0u>_;O=h0-Q7cQ2=4A~!5tFZ-CYKE
z2yXYwJ~?v#hk2dp>F%mpRcm#XIJ9AgElN7(IYOI*%GJEvTkJ#TgGT$DQ_Ce^%}>Yf
zV(=%JYId$hkUz(jJ7PAe&4()0F$DmH!TTK3qa2lt+~r@p5jwNswJK#wr!5#Ns@x@?
zhPvI}5V7oXw$1)%mOZ-uT<xq~#CZ>R5Np|0`nAN;s%TCBV0uPHl$r9iq(i4kS-F+|
z?cty|{&I7_(reRQf6l1Y?SL}cH=Gu&(`Uj;h|9BbIu2shoaGAm@*SFC71$b&<z7q$
z5>%k~<Ua*1W}3;5BFzJ%yZ{c7t5bwN5~knx#T+5LOl~uApo;dY0<!pO*RIzE0izo%
zn^p)vXiIK6FQ;*B{<=8_iZi%j>!}-~Mfln5eHZBMD*N)w<!8IoVsod4j*DUo^FfL8
z7W`Spla5N|6<IwqT%?kl<M)g)yt}KZfOO=-OlKG<R$UXlXx7|85Y|ByR;+vFPe<;-
zM}Oa_(0kI9yPt?pc!#Os-i%UC7h?%++2EJmtmbnC8j4D_%wEhBI>BUmPIb|G!O<%}
zEoeBwv6f|erE!DQTcoyiy+yuBYTqdoLQp#a$?<V6yX9Z0f`3sxW9d+EX~%9gJ>)Sk
z{vtC3i)IL?Q(?&qR3=PN<s0Y7YK7J`b^V*AjwTfxf316I4HbOL7<}sH9XTt=!f((j
zEc0l-pi3W?>~pPDv$&ZT2I@+Vj{KKm9BenHgYu8N8P%iB3^TL|tx=l(_kmrqscTUP
z#K~mw#P(!Q(OT}&OSO5lg#<?NP4>YB-#XZ}7!~if%`vG3{GuN{x9QR-4&rUyQd;%+
z=>Rxgih}dmU+Ccf)Il^G4mTtGOiMFApIv8G$%GZ%g1E*lQSlSEy>vgg+0SL5G|0Yf
z5Fa|wxz(|*ZhSa3Hjy*K3Nzy|(|fcbaFaQ=$ZQ1)S3J=ERYC;&>^h@#j`sU;3)@#z
z1&(j*7l~JqM9X_=(Ttl!?L2rn1JSY2p2RiuVUJAb(C`wP@V$g?XRg<s<u9hU@dGPn
zB~R`)b-h4_H-}wt1mDX$o@mDI`8(QK4Z*gi4!;*UYOVLFZwctDMiVQkU8S(6y`#rf
zeWL*6NoGHEd}cYt2{8?#k|8d)d<$1tZt2#0eJV{pLtuS$eRpG#kyXceFk^%xsn>CL
zmL%N_LcnA2Z5Qr8=1+3qpVRDaA>3ow7?SN>a)vg{Y-<s{DqpKxgq`)TZSpd?W(Ho5
ziEaL#!Q558zUaK3XL|Wy8?`N=6E$<X6)<8mxa;&d;%CA?FYv-=qGZh>i6wXE=Z+PM
zaL$vjP{~I;16Tn6ZAi7f0GYSm!f|#v2l+y_mAme>NL8iQVA~txg3nV}6s~0}!LjSr
zFiG+X&~$M<KWb*aeD``!?=Y#4I0m6BsZxJ~&*Bxn@(>tM4hJ;v`BRinxyLy0%G(OH
z2kP&GJ%qKp_$X{dHQ`Ugmc;6SgLf*nF+7^@2+4RfI_}x#Si3S}G278wE6rxu7Wf_Y
z5+&Nw))MWm28l*>HsaWIy2L1LX8N#RTShl&BS`6V9^+Nm&giyLZIQHS;<?r6<J#d(
z4%DL7UX_8bNjq@*g`1k*q5MLb<yF7amtSw27n|?HG|5bdTGutZWcJz*yIkO~p}UTf
ze?s0w`h>`&WK~UvIzqt-JrbQ3N%ydisILr6>$G~u`MTCsVDH;-shPUycg0COt-Z$u
z%(i@d{1d_w&0aj*KtFtGnVv0pCC_9K+}oSe1(;p%kEw}~t8$;qLw~<?6(Xx=sVS)g
zz@JXPT(vAT!!e&cUGjA=;IY%#ZgJy_eO7A%-Oso62Wj{wu=C_jhbr%CZhYG(X3ffi
z<qsz)EtFc^Kny~}UUwP~*49ozIL~UB*5=Blop2ojK~~OR02P+El=ic<^ZL^In|KDp
zS1|#oj1r=TSA>WM1ucxwt+AUF-QJL}>yq@HLV1@_W0zS$aYPKbK!N{{6DWgNEnIP#
ztZ^_q45ZY4OKtOS3x+h3Y7S{KiSC%?OJ>ubZcb<Cvqj%e1t5Xj(wZ&GozH_Hu<6Je
zDiuX`tE!edit0Id49e5<XR_K&(N0$%uB4F91S{M;D~&Nv)e#5?wDUI9i~~8DSqxas
zHZ`=GC%6RDk=vBJf|Lfo7xLbEJ+-}lmzy1?IsWCDskqJnGwShk&rmoevm;(N^qPKy
z#KwH5XfkO;L9z>4DLYk@xA|AgRI@xEbx|Pie1DNf)9FVlM-Ek@S&=gyr-3wlWd=U8
z9q_7vng>(kvwO2Fg3j+ZECyi3)VqUw#=}t6r!o_(wnKZ-fT?sxxF(d%ZED2S#hZNh
zIx-<-Dw;h55Ok+8Qd7%pOuz)0D@s+1FUtY^tc#ezuOwl>wf)X&^GDcsWj#Z%Gu;O-
z{V7!Vx|3wR>%*Ls-;JzXJA#!;5oFIHkbkc-JEFJ|G<?|p)Oso-yHS+|@*Y26;=*gy
zSdq;~zoPYSOjhMuus_!G0*e2B`!*mo+C_S%Fudg2L{_JV)-*>_@0>YH`q8yMVzzZ?
zfx@}5L**~58SyLO^56Xu^Lx^F`s)cud_;ZRQoe?PWl;au0Acgc{xQJBZ-w7@FIQuE
zDAp96%x1vuM7UAh#=mR2L+gzQW&-mhm)Swm`L*vDgoT<3#`h|}Cz*e$3~N+m7z@-3
z?c1e4*iA^5%q8Q#iG19znGLZb{6RD8JigaelF_|;LE|!`zlQ#U;LUpXC!HE!U3jM=
z@>_l$T?|j@PlGm<qKxi2_yoWE+0W0MW~)9vF8p;4`H!}b>dh71sv7Mc=$;SY2xhaE
z2VtI1DJ(Q4&@ei^P&zdk(_kgufB%V)KkA<c=7P}EN}n_g6-gK{!4!1=f>9+(j1myG
zlt-B9s@P6-PKorcAPv)WV<QO9dQ3A_aWbZ8gZAb!I<-j}si}#40rgnxpSO?D5P!IR
z6@V4|$62Ldm>W}z+485kBOwU}1ZPyJx(w1>?Kh8D+$?}_%!hTBlWW&=bu(qdKEYRM
zfcmr&>2v4pm&fY!16(Rmf$m)-T3s&}sJogGTMp?E0&|evAr`pH`CO?cf?>ep#@4?;
z<edZb@?+cgdc@n|Xl9sfkr`WNZm}UIwsSo^m#AlPc|^U0d?CvpaFksluwI(#h`YeQ
z?lJoOZ1)qER_I<~euf3;CkFnP`N)Q7Ai=|QjC~=EY*8rVJk%O~PziQxY$`51pVa8%
zUu0I}AQ>QGxb6-yKPZjUd#<Yv+yoR&byf<J(#spd5pnu7dwy)&)|$PqjdgV{eF{;j
zZc+{0z_tJ)8yBD{{vQA<84C0xP`3baD|Or#o7-_-eW#d!Yk6RJ5M|&$MdFziKMPHF
zz(?ZMT}ht8R+u%ZFN{=^m_osLN?Da12xGM?9&I`lV%qrNuIo{-&m!fX#<G#;aW@lK
zZhj=2@EhXvX(f3~&>1?-Pw@0%B>9X;E@(rVwB5Y_bXeCz%JUQXd2kk*^7M7cpNFRZ
z!&?4X*6<@s8<=|Z7-=Q4jcsM*!{X!2YgT<V?9yWx#;&iJ5z~~bNc6ToMKPb0bHn1)
zW;}(%APw6#y_Q1y?Q^Nk=Xq5b)88I>sw`2?ajTx3&xmOgONh2cVL%pe6MSO<YN-vK
zi=Znl+Q+#%XJ}hQ$Nv|*`fsjtTRDd~Oe{p_<Xp7QYFGsBgrb7ch%twTQ|V;KF<{ky
zz0>;k(^c^+f%LBhGNSQZmDz6}Hb{=;I>OZagtdK(d-pA7J=lJ;NqJ==zt9~6s3MB(
zTwr-NT+L5f*)}H9v9a)cdaJV&2^VPoWnebw*RNXy{pNR{s<b?tZW3{ZrsW$C;c}gb
zuJnJnTGA?^tXP&5)~?3;Xf65MRD<G?cn<WcxVFvjWNd9oN#gx}2??O&dNap)4oq@G
z{H3R+)yRID89U{yX=Z9v{E@v43Y2}F%hDpcI?L&^DlLmY%_zRuQ^seZY+9U{hLC&f
zVF064{y&0P0Qxb^T!uQ>ii_}!YKQyX#1N#Q05*>#jP~?Kf|>tp2EaF#7LInPx4qen
z2C;;33@l2USc}UpUA2LRUUZa6PK(%0`Xajq>x4774z-roKS|d^wJ8J5J6Cqg`UM#{
zPc+s3F$drl{F{J|pHwGd*9f$RJWzl-J(n%-EO<aMpCOfLiEVV7y}GB1$=W6kkWVUI
z;fk&jlGeLkQjw&KC6>!@zlp^(S3{z))R$-AdaFgLH;WI>{B+z!wE%bAO8S3}`Ta0_
zeI3G>mr$^6(YnaW^3!ytu!#=!BV*}pj51sN7PYuPeCH)iIMKL4+yEOqh_Hs67TS<_
zWo9k;3|BOYl;xH6Sw=HWQQfkf3F-3xFzE3RqQPp-Vv~8|yPXZcs74Cl%JJiEath8*
zXyw5pTO`W(k93w-#YXIuNGK`KXmCTe5)$>}j7kcAtk^47s3H(6wo^_Q3!SL6z0RuX
zJP}5ADM_}kB%8l0hrq?cY>M-`4W?$I(Rw6GME=Ntv|G{&58{)^64KxJmlgr>ulEbA
zg>b024<y5!!}Tm11##Re)nLoT(k#d&Ip3!0`dOs9Kx@sYWCUhXog<tE*AFesw3^nx
z5vQWbWs(>yg`_DQ!uM7sUbKUj9z!tLWju_jUl>df2XyGzouQ{q1C_Jbm5hEvFH37J
zO-gCSo7IDu?#1@LY!<$o7?yGC&-)Avo`tXp&;6$8?4EvGEGG&WAWb=fjWlJGuJj*e
zL^s%fNbAQ`bL+t%AL_zR#;bg_>GwNTIotX}{D9CD#C{5%JGCDX|HF~sNV-NoG!KzG
zNGy|yr$A1fzZ~ZD;4{(L@94tS$Mt>Jgs{E)HzUT`BAFBiSes*sEM~2Tq%Uqtm?aq!
zp0UWkr>^$huP3Tyy}a&+-<776uIddc?qY-@nN5kaz1h%a$&dGFkvWM#FDHGV>u1mL
zu?pqWA_EqR3iDq_zx@NL=vgceDHeOE&9k}egf_mq_}<L8xu1PH*-05c+(V|0WV}5O
zIGTj?$)rv?biH!7-kLxPKC0LJi1<H#dgvZ8nVwv3fBIX9xs{hb#~X0h3W7NY8SI}t
z9}wZ;F;w_$wb&E-*J!)r*R{OGI+&^mpdE=4e!p{@{849qgX2gvJq_&N*9%}I&_B-n
zfDSdpiHnd(2==wmr$dzX_O2)ZO@$|PHD7i~BSgdf_Kn<wtIj91i1;}>z{iAvea%!4
z0CgTHgCIv87P}yXEe5UOCMaw=)gw%jS-j<tG_jc9&moV5xgc~LRdxRqKyo%-&h}dd
z&p}|!J8oW<Y`~ud(|ipo?KK5N;LS2tDeQ&t5ekr`kgWcV-+&v{56$al{tjPX9O7cp
zESjiB;zDA6;-d6MT0C_yLUYJFd@8KSnTkQ}2Sw>CZB)ImSH7vik0&hj=Vts{&Y+N^
zbDr={fh3ByQs_Ges&i{2cF#m1GtgS13}VbrTS+4_I5WS4HCZ_L>lO1q3qbr^V+0nE
zBmy*dam07WEa5mkw9v(~Rh+{d<U%b7%HvTc&S~#tD=p%P&0#i%JIHC|DFKt7KRIPL
z6W%{O?jp#zQ>#7!D)Ad{+ISF(vF-rBZ{^2i#bvxZp;!*vmjEE&Cf>_qNMIV<-#e^8
z@?_65nSA>q9Rd@_JKf5*npZ}LYJ>}?WHe~Oc4hdmkTU2`ngd@)?WX=zdJH7U9=dhC
z5uH(OIaAv}u9>AMxR)@2m8pXQs6I0uAM$372QpqvPCWGc^Z<Pd`vxW7yuc1n8_%zE
zC1`AGP(}h%2>TxvSrO2g{S;DIs&Qa!$y^T4K}@PxdIIR_1x|V6(=Emd&krSls4FGC
zQ@*)Flng@gS96gCpH-bNeCFd)-ZegBXP%_@+WN?T^`eE7P#{=|%75XedfvNG`{-0R
zt9{`)Rgw>zLNR`$mBLT+mgQfCtXHh;_zdR&&T{;jY_GsG?YjL|{c&SPeN2)`W{hrJ
zN43`arpOv9uRB&%A+Dy(SWRn7aO&aa$wc?kW_4(|v~PbgQFc51yGDb7St<TbEtm?{
z|0FOfD9~k_La4!ChfM%3THb9KRGv>O*x)BkRQu&7Zgo5Hm=inEdac%F#T6_HQ^1cS
z!J2`3xm%3Qdn(P0ryxsFKREc_+pFq$y5Ji>GBz9B*WtfL6DIGkg(mws&<pqp!&(0v
zQWF#?1Go|$Q~EW%r@Q)@gq3w$bF2-c$FlFLlzr)&{G0d6?H<}^tz&CNmF+JBy~+ad
zAwjlpG?YA6u(silIq(0cjGDhp9m9i);ZfPh`wAR-ZJ5-OQu$n*aw058mUT7{RYysQ
z@iE6D3fzaFKJ9o%4uN^wcM*3(H&|=W!td$fzDoViBGF~VfJNd%8yb@wB>sf_gabjF
z4pfbSI*Br|h8>%^rsH`37|s)TB1E>i^i?=M(y+k(NJ{)ML;K;=eZsd)Ic(qyN&ZrI
zf8;1)E=nZ0aCq<CN#KNqOn&%2Z58vHmECpCMGQ)5Ya+gYC%((JE01XO=GCA-DDXre
zeqwO7_+?s*z|RWy4f=odIdhW!DPRA!phw8-*%5;^qSPVozLaaQ4Adk{YfqgP^0WF0
ztj;-yv*8AO-5(@i-~O#J%Z7LnW5Yx{W3CH!6#z~Eo(naQG>0Xn#rcVWAx^HW>EhA!
z(rG$kQ*cw{SFxVVe6V(>JXgzlHv-=1F|)Cffg&34ufBl&i;x7l(@f5HmHbHTJre+>
zebL3^F~FBi=aJ%<evNs<$38s~#a|xG>me_3R!Lp@hV}%j;J~}I+Gq<3t!1nnEir}f
z)qwf-5lV%FSQL~oCsY-LR(BO4)o9tF2qp)W-eh~s<**BlWF-ISyoJC(g+j<4h0aej
znt=4v0TAE)@xZ`kT1J)SJzBg$DeMwgwksTru=pMH-iRqNy_7{T4>F)KRbJ&sPxA-M
zw~gr_)A<C?@#YBE-?Q)CZi|}t%~unb887lOg2^C{-rPLv2sgfrOj|XSy_4;f<Y>Cr
zT;}|lTzr-PNE2i7&ynSzfOo92^ZS^*+w+5@%D}YZTw4b^YkliXxjWt<&oi4?YO*5+
zwr2<7UxyX{mM4Voi@k#Dz=0&7NhcCXS}d{sP25CeuvTh!pjDsx(nOK{W}<S<e9MS!
z*EH?!2b06$0Nb7+p7Z-ip6CyH;@3oZiL&NfZxk!u!i_sP$G%Bgho^OqUTHqr){KYj
z%YuW+m46n9n*WaxS*Jyhcjasu3iJl-RjxO`6BgR+t#etqm0S1EX#$j0#cweQ2^i^P
zMVYl=Uw8f{KJfslVY5cv`T>bsRokb43}rRU>_JgEI7==fD~I2wl1@!)QB1l2iKUVC
zL2P+lSK3)N;b|2axlovVkY7?-#v@ZKgWSawdl{WH;zSX=mN1!`kYUGEsXX0W9ED*D
zO*ABUWKT*w)~lAuX;-{K<xe>3CRAw$1B~FZ>-LapLHl0l7zD~NL<ILo{X`7cx|Mz)
z_=d}8Ia{*U=4M0!Y$)7+a<msYwjFA+)P_$rnS;{Q4@k@NnaGIw0X-I8^0>n|E}isv
zuC0j0T>a9oF9D^^{FtvFxaTmUbl;|f8{-CxK-*PaxzCyFqW#a78sW8~(?#P~`&){g
ze7tYD?RI^`Kn@8Ypk1Vrf&kDgYq?*UsBDL9twYevM)6Ihxx64Vv3_Gsr{P=-Z1%85
z0gPSdin8T>Q|Xmbxt)1;gH&Zz0j9o1&<QbUjeK12SF%2mkHsO*F`3-btYkT&6@Ibx
zgYqj#_eZ@MbiNiXs*^jNO^RqCZ@r*@YN~Y8#)S_V@kaEUZmZ$ghsvjSKf_Fq8&9j>
z5`D#9M9mhmHyQ^fUteG;sHpyTPvVV~b98+`q_}4TD7*oVcr?P4$_=7f7Akm>LoVi^
z7NwJRadS6)NhWhSWxLn`UBm3$j+chfRN|?7D+-50D(6AEkH>z69w<t)m~4Awn!VwD
z;k9L@I1y>2UAVG};abo1g&o2(XE7^__Y<W})3p<YEE!Zg!ZZehG)l0qZa~ttW4k=1
zjYz5mD0Lo`87xZ#l3c3jhaXjra`;(PgtIV|nnr3KVbh!hBRgcUF2!05m)iScJAr1<
z)2quEE-lqpGq&-SCAGcbCZ(8kup{eb=jK;qaekLh^>#7UY*^`{^3(U_4b$#xDa966
zZrSVBXSBTC5zPcu98Kw3O<@a4N`%XTz3M8^0e4z9;g&^0M4T;ByZWf1j57#KFC0<P
z=P?h>2)0j89V~WF%$Iw_@3}l>^cy-gtYTVlMc=HF^NP9_DH*(OYTCAm(H*T`(t3AC
z<(T-kfsXVUwO2%x*Ne?Aed#^6fmNr?tJhs|URy^^(a%liUlsGkY*rePAV1B;H79zt
zieme4vb>$Hj}Tj?>$S<^b>mDp4?h>^%k6BH11gV?b`NsNPxs#N{|WIWgo@6hMq?Ax
zO#2OsJdHOb6Ear8E?v`sRunx2_!N-EARoF8C%M!+)YG$K6YVRGNR=}oZKY)7S_u(p
z_M$a8VR>bukrusRhs9c$Z2DuY22$(@HMU*fd`$L7h*+$3R65w|?j<s&qR*5owU66W
z?TltSe%vS{i&b#6hxDr3B8(o|@>(zCW4>R5vngo^cB<mM9@N+h8oKUlZO%*QlA3t8
zv|8@$xUk(TEBdrS@ITyYZXAO18fz7?$`tq0X^CA@Pp}tgk6A64^((6iAEKqKeu&l5
zJ8VCTo1Y8WIEi-A$vU+174%kCp6&dY5q;ornNQ1WCYNoNQv;cTBBR=rZdEuu-y;bs
z<K-sIjd&^0zMoU|7izg+zv0{cl=F}(YT%2lv)($^^p-M6XnPa*=A(1AjHEp)x`Cqa
z;455Nec`=7yZdI)e6psOYUYFHq?q65xRRTs_Tii+m3uqc4zVYnmBHl2WMg>f!SndC
zgA4{4=OYOn-N=@``95RG;|Y_COG@kOlpsr$MUws0f&)NidG%v67kf7zfIQ)p0qI{J
zpn3gKqx0;tsw#{AN3wLP*8Fne_x$NEUKpyOnH>HwO}^i-H?Icc@ytSiQ?;hl*YRBG
zCoKhyX}*NbQo;O!j0VOSNWN!LDW82wN=QWR;S9a4Cs5sdt-lOR4JobEh~Z4P$;dMp
z=KFGjLQfSU7H^fN>Q8`f^4lW#?u<#TEr5JbAS;{~2csm}+S$s@0i~uw+T28WnKyJE
zXuT595OVy}qgLQgTh>xF->G=5T~@)Oo>-;w1_qPOh=pxG&R5a~M|og|$y3O5#glPC
zsJgsQvg>W`OAqx8A)bprMeTU$QaG=L%cd;7_h;H6{765^HgSh3^m@A=#^vt7BZIgm
zPm^o_F;ClJIZ<gJy9d*uqc7;LslfEk0!0m);`{OJ-CPXzmwdmZTKdv*0!VhPVUX9|
zYg&e?y>WEei#EI)EMS8U7QSIg7;+T!8}26P++_04Y!BxG0grd*f*w~O;SgjC;9Ry5
zqZmpbj^hyU8TR6JUhEw*R!H`S`yHAZB>o`jZX9*?u$^k;hD5m+Z?q=-CcWq-d{-iS
zf;!kw|Dq|mIiQEFDrjP+%)W1>lFD`CMydH*D_#-12AnWaU#hlP*vxuNckT1hgE`h^
zeDc@OHAb`y7I_TGgYCD*T2XAm`50>wpv)!313g$ot1!NR;Za&Z?-_pk!*`To0b^;t
zeua|5HO2t9;c7L`a#xc|%#$wrHfu2|VO0_)v;cIxE63;J^a@{p#EVznctP&CLV8X@
zRr~2O`$$7!%dS=wAPjAF%<i0rk~Va-)ur$YZ1TPV@^NV*+w}hAxsa8Df=c>%o~$jY
zw#ybk4({bLb3Y@Hj0_OiU6ErgC5k7a>rq<egBk2nJP``%C9B1}1?#u@t0Z?~ci(-T
z;D3AVtGuSQUZqCG<=)y22}Ww6VK$vDpKrIs+<BRm5wUZ2lx+PxY4eM1Tbt#B7G0i1
zFJC|!qr5&|n|68&a8}iArkUOI+<dZFd&7jr6Y##T(rvS6(or?@D?OfhZBb_nQ}@&S
za_Uhky$oa)*DMkD5!9?*6y-U_n!i2-GB<0Q(rf;~@$!%K9gP-zes<Y9c2j-&13v8T
zySFa`=AZuN8#3^0<$1CDgwWI8Vv-=_R;#qzw@QBV;mWvA$|QP)HR}FEMQbk&)&kI;
zA;`5&qj6qy+j<3;>S26{g34li!P&X|eDRM}hG=!v3+Bq}c^ZoSwBfaz%zb30ijvp?
zr$73smmof*VMvUbDw0%L>Yg5cs4CM!LS8p<B|_dez6{CKdB6tQ@+j{n;`Yp}7YiHJ
z$N9kjj#i7hPb?FPrk&1f{JycrwOBE`a?2tZo4yK|t>@Xj`bdvJ5wf3br)F^UZx_1v
z3OPRj-oDh^UjHNxD{BW)OyE4-WjvYru0cwm?Gs{CSeXz#Ey`P#<B?ee(F1=_vg70P
zXkO(C=rE^vxwsauhN<*!JQJ|3I6JJS?3gCk{JoEnFbr+Ak{N0KH;2#FhUd>Tm){@L
z#2vOE_c2#*oXD`tUz^MKwi>PwMBRhwHGEne@ipqUM%2AJs4Fy@6qR$dJ0o9@vwu|x
zLBb#oPE5G*Ft-_D7JV7XE#@)FdMV`F?Yp|x6jAHGifpm^C!upJDBkJL7}IWYpsS=^
zCFt@t+96$fe(A9p#JhG6`Cifea+d#Q@!u@z^~Y`SNLV+=^!{E$zkWV0@<5sCl_UOH
z+c`09Qy@@sW07f7K0Z)a@Q>S2Xnmy$`$XAAgCFiA)>cXP6@p*J3&vdZfr=;yVvUK0
zKmeM>SQ#Mp^5rrZcDTbw@5Q&Rw04Wv#c1plx6WCN3-CsOxpr}G-TZorwr|(A-N~>1
z1y2~E($};KKYQT;x&QO#$a+VmsK4GBJoNJG;JGv~M({kzSNR(1RsO|SlWxK8jD`Cg
za9-IrqN6K*W;&A!{oj&6L-tEv9W1KKp7~z=8Q=2&<I%8luT1ZfAJm{^cvzTCJc^`~
zZ`5a@tGqZrEU$W!>rV#y-2bSj2HsPIEEfP(1F~1Q#cuW)rFp^MUhrm{W2NMdf>F~%
zLPxyCUxX#O##Mg3Y#pbEL9CUGMvTGt8XFrgp1PhQ!l*{R6jo937-#9}3-C61XQN^^
zB?+3OM7)^;6FH9Z80wS|tQur^TC!HL?-W*ME4@y|kA?o)jW_&kmbskh&QIzzP3d_&
zNBF)jnWqqG!*06iwfz*wh|wet4cwPdudrJh>FYH2^Akg8ypqHnd@YY;<fz<Aeo-O~
z(nVPKl{g!SmLlchl{BwVh#JJvT%ENdBE>UF<?t(c`a`LB+XM^_f18d5YjAWQ_}RV~
zS=F3l3*Lxf0x2K)epfD0;}1$B{k(dsnq{qKA8j6$U3F;vO4*83d|@a%wlIrP>KokA
zb&9lV@57Yj#yOXDUGszjC@FW0PCGk6O$8>i)&xp9c|!E5g*&Cq5x-m?%{`i>w6>8t
zd*OpUH30g=0MbqT%%&p{g%RUVfGIjhZ>G_T8uBRXiDT^w^(ZzI8()@{f35+|PGEJ+
zvY+-bTI_q{I-OS!tttfURd*JZ9w1TG4p}d?+Vm7vc6}dj`JqU!t&pt=ElR0PY(+)4
zK22ueYl>#JnZN`~&#R6r-r8#6y*wOuE?)7a*svvBj*fo4-R3&U=9CIFG|_s4RV#pr
zW#?0D_CY<Mkj_DrYdk1tyIRP>!65~&?dVFfi$I|KA!U7rg~s5aKkg!Z|AMR5<G8UP
zBQm-nal>};+xl^;R_*mqMI`9|MyjjPyux({BG;?~uR`&x>&gQpw@Fdo1lTW&#vHCB
zZuJ#;bdx7f7V3VL8O_+c4t{bfn4t`Y)3R6lfmVs6Y20M*uiuJgwFou*VJ!Ywls*vq
zEUEQtFp#UJo6F6Ai`E0Sv>=rb!1}}2T79+1`5Rm4<E@5bPL0%%81&U+n52KOgp@Br
zDnz=n>FTLfRi;k^4X#+GIHljP;sp)rid<mQSb@;`zs&3_#37<?GZ2lhKrbnMR{BqL
znO+tyUdGO832up3T(3WZ%83<&&DRgI#`R|1C2l?RB^Pq{zdwIIzaPC2wW?B(x;OIH
z>*9t!3eUdLZFTIp#hCzNNU|?EXCwcL<X(bTt8f`2{G7JCZry!);$fo)jhhLVFc1ij
zIc!TWBUl_hmxD|&Q@C<lft^RdqNiqHm>d71P>4gI?Rd%QKt;17NPpXWeA?AxlM<&r
zD&XPR5oG35Q)1C|4+)E6HKq$Zp8PATD1m%)z^Jj~V(?50&ZIeZWb*6fzT&}~Cv~|H
zkL;V6=`_%KWoKycs29&<YU;pzriR3uK36xk81i!d80U%-hrfR>&-igmb@)w>z@nZ7
zr#tpBizS97*>S1pdt-jWd3JLHwrr=UQooMOlK#}PzUb#g0&n<U+x&`S<(w+5aN`-w
zP#;sqQ{qo`XNj<0i%Z*jd2?Np;bM7~0rGhc>N9(1WV9@Sg(?Cy$h&Vs5&DyQXd|TX
z7?isbl^A0?S`2p74Lk&tx8_PYEufUo%oBr+!C7KX#CfWM68`O0M>a02^iSeajoVdh
zpD=0XToZ=1J0v74z5r+rg*s;Ze8-W=lER4Pa)h&Uj$BwZ=Sd6bYwb4`&{a4pMc3XS
z3gNcaiQm|sV?V;n-w;+A>7?qGS$h>%-DkfXHDV*+{SXvcV0U?5@k~}t-VhdNKqhBB
zDQGqKIOg^YO8>g@-cMGSo5taVefYPtoh!;r3#!U9rulNcg?I+W+1IhY4RyZ>M$3=X
z=cjs0C_DVDzL(C$qtQ=$-hjLiP;R_OxaHFMlx<*mO|()eC?%hXaxgmkneW!AeKR;C
zS<!ikGveV1G#z{`E}K$gG!O+Zhx@nW21>ykWYzvo`D$`1*ZB39Uj62~<txq>(>S9t
zYH&JqUdZC;R-x5k_OEUC8Oo|8Zfqv==!x4HXU9%N;}hEj>yMbg$W0Ys4*UY5)_nns
z^6e1Yl{c=+()5lw#$dXvD#`i=65(#8fc2-Ii3F)AKKV09V;2MN8u_GQGw-~2>D5;Y
z=F{(=G9)mPdm6Sm6s`Kc^)?vNT&a^ptwEYV$<QA-(?kmUh);G?8FLNtOJ4z90sLU#
z3$A|&rkb&b8cay)YhE?x2{pZ>xXT6LoT)meH(K83u^)Laz0q>Eltwu}MTZ?33DRyL
z&j$xTj?)18V4|Hc-NsLD1=X-7<(%E6dd}JiE$3&2j23k>2+1H!e52D~h7QvzrI7El
zVd(W_-6($0X*@2L04;cRy+7LRrNz{;*A^Jj6~+4`-EQ{5KBK7JY;Tpx2$&du20W4k
z<Sr*fy7eyE{O<9m4e9M?6);!1_x1AgJm(3do6SEwMjFq98Pti{z3=Ycp;O88dpCbJ
z`Zvr2q-|V_@KOUK@gIlmhEHvj?czfW_GXic-QqT4(B5+dD(+7$t4*KpKYPK<OnYMZ
zL>p`WPADkuq~r9w%pSp?+H!~+;0SZX8)=43#`38~%(6ZashN<<@`#=u_l+|fV`6J-
zETHJ9;xjK$`S68oMRXdU?S)ZN{{rU}kA$WI)zZQ-AT^GmCaS=Gu+m{2Q&Xkh8nRww
z9l<qvRK6G89Ss}*js>%kd_`oJk)`Tcq@~a%O78(>H>+M$-fDLcUBPmTpXSm<zjBd6
z{f*wvd{?lg@#q2>mh(K9diY6B#CDmmHX+j7gmsB`&Tlra8OjAMIiG6$PWnH#*3&hD
z?|$SgtzTmrUnW#c>sGpU6|h<}GtVnl&!wHPm+zOHwMMe{Ex3lJWm#tX>MSQ4wDu`O
zkh{2fHwCi{;^s06Vz+Xoq7xK?5H^xjLw>`(=cP?)h6!c<fJLre;zTOaSB<p4Pf1%P
z)cUOR+a-f%HV4M*WNd?Kb)F2(Agnz79xr^tm80KFtJoDG4Q_4qv+YUy_DXYoVxPg2
zwf2RCM(3rY3CTf`T|W6EidMV3gD5@Re`>8f@5cm_>_wFiDGn=$LlZXN7C%T2NBM4p
zT~<$ygM(73xcQw11%i;H60Ji9;+={>DsT0&v4xiYQ5MeT85BgIyrLK@EW)gum!Y=>
z7NEJH^PP2%mT%ZIaaCCBJI44if2(q=@wc)<hmu{V3ng~SFC54_j2B&n=!vQc9m?cF
zy8~8!N4pwU?Ar6&h|g5mCF*)e1I{#CIm?YZAbFIwaA0U^*0!h#mD)$()jBMOglING
z?&0LgbN85Ru;mUnf#7!yJ%xHvVsfPrA0GMfw)^A0)$H;Wa#q)+;O1TYjYn-3dtbHW
zIZ^G`%tXIbmh2l=LS<}I$>3Ig);aEmmg7>aD+}Vw65yrJqtWN})<|tn@%@bro>Zix
z6;dTwA4aIg3i~rgH3k31lbSi=6gLl}Q-0P8e&V+_XX{k}Dn_nu?K!3MqZ@>cs$ONK
zX0fUP<^NU{?riz0rcmgs2!9XCHk*NkZ}(RYpU59X+85|%uFmK$-cB;1H3)|0gWuXb
z@2bw%T{ha%ZU(jHIgc4NBih<f-Q+`szLSuWg1?VLr6*l%e%h=G5DCQxe|)nn!wvRF
zl}Pz0<TK&nUQ}K#r|yQbwx3yxrRX5J6mpGm|GF_tWx5Q%!JyrK<LRTSis=W*NFFRy
zMIS6QKHES(xr-RfgOlScGclUT@az&SL@0~sU*c`G_gMoY4g${s;p5CF-sq_JefU?>
zBljG66WFm9h%th6U7#KhW0ruG#{oeOt@}lDGy#%tP&B2=5ix?rISHm<{d^o{9FX3=
zO&KyJU{srVZl*Lb`|G6d<fG9y>@pF+#S2&mFP%u6T@G4MbBah}NY7_ioK%c6Q`ez<
z`yW#S`=4hsw#hh_t`d8zS>u)IKkh}}vd3OhGzUYXcj^?Wn`E}D%o*a;yZfE_Rdj4}
z5XzX<{p6OU`>bqN!^k`D<eIY|EDH_}u1>%{fI67ihdP*ACYlpOfIU)?k1w1PMHMWh
z65ka6Q0hQv=xn?+W5Ijxf$(}cNoj_?vK)HA22(GQA>g<9;YHa<WE1NIqDE%f6y|La
z#v??+sZ&G7G^uC3iWT(2dF0P?e}Wa0VnMhPaKNgT{B}({<|7Qv`8%q^Ep#)zunfm0
ztFXYjnm$C1dci4jo+DRhU=|aNy6(62f@WEIh1=E~%`~coE6*p1PA+BiAiAVj^;+qS
zmw7f^#yy6YZ3ors=q$U^#vDQZF=qx668<)Bx%Wi+O`619*>dkLHM-ul8VaKdlc-|}
zlbD5*qOy7!!OdRwqK48kG-)J$-XqQFk|z&V)8_Oi4lP267^$}qZIso50t--THkUev
zf`K~$<G79Gl%h1fw2j2YGKL}9Vug^^=HLQ1Bqfx);G~JT%huK?t1<0g=aJ*h!V}M#
z!KQt(mdevb{Ux}m1V*7BSNhO>r=O+)W(W(tHp_XokT{$%t*px6|A{~8X!9a>tNZur
zo@^j$)u1kF93T5nK9ja3CdRr!_Ot3qjgLlVWM%d2jMuF`yQ8UCcRy4-Y3z0Q`0okX
zyM=?R7M}Z4UXJJO^4{o^8FjbMKM5Pcfrf#P3^SCWsU4y}PZkaYb4xR@EXu7G6h!Qo
za<>y!qrRHfR>Wav)o#iJN-QP#)=IxsXFJ7*Ay(N8gBG+7K2cxc^XBmxllph{G!W0v
zomx-#=2IQQ@6U)b>=!lzVM7pF#}d2<hs_yKKPOVV)2}3d8_tpX^4<#yzI<qM8{+M^
zM#ba#NAL~|`AEza)0L)o;xNXq<?1Ks!sH8DXoDjo2#1lY0)MScg#m(LyP3<Ba195-
zg%4P`qXYXlq0~Y%{S5lHUp%j7O%&-3Z}kbNlIQWUCZ#`<<dLIqQMfg<)Li@gZz7K$
z!dTqByo%bxSlg>3WV}o>QoiTZb^2Jl8f7=b1|tl9%{%sL3o3%RSd8nSsn2!a`Sn|H
z-qO!$VY23%O2l1}cZ-OLHeHKJgP|X7rqWv&b6&Q(MW=QZ;}16b9C0yJC{Js+XrY)T
zYg^DdGnAtV@F@azwT=^T2J(phDLOilB$MAJGumL{CHwmyQq$HOf6ii%PM+`Pb~Wg)
zMs^taW?{as_u42`F0LwMDA9z&s2VI(9^Kjd_H8{o|K22VFaBY;4Te&I%+n#S*L7eY
zib{6L_V&9}zCDCmDy~Fp$-|P1BCDC^d|+{_W?KwtF-PR6=4VWb;jG305Ue7}ilr5Y
z_8rHpace`5b}`NL3QX@?i`p^1v@64MJfC^I@0F{Qt-;|oKaDcvs9*KWi(AU(!@Iak
zZP}6A!k<G|J~X!(jk^lL={G`Yvw0)3Gh3?h1FW=bY99VIU@rIWopH##&&K|a(;RiA
zJ<~byk8l);Qgn1o5uAwrI+}?78KqiKTB8o{MtNF;x~7!ioNo<HhAt6mf1}rg4RqIs
z+?9@`=n~KDXAKK_c5W*v`&FC3^@6r7q5SpI#frduO)E6ARLmpHQ?;Z%8@=x2OHlO`
zhlp3`bx{fkypE=ppQA0%Lp@p~!u$?;dhHe&LN-VuU5m@mwH~hzq;bk7qU)E1MYN8X
zpl76k{K>~7<$kvrh_>GEZFN2J2ziJSO+n-2Mv1MJYp!d&oIYb*kARCvyq?Awp>Sz*
z^=LCmzPseGo%yX|d}fzzC4d|XK{)Q03c1gBwxI0+LdYF;O|MRVe!0Nj3GC8@p`hsK
z57}=<x?&G|oepcJ!-6Tr&2Hup(>61HkX3qF^c`jr-79q<q-`Fs(h<amfF3KlFsq<C
zVaiV;iv{TAKW&Lo-v+c8_!}O}qJ>~lh>YQ8#r}F<%BZQ2b;#xn{z7Rd@grnc0{VGg
z25~2s%y9!^5yZHZzU~R3OE)!C15#PIOfQb9zL&471>H01#-zbQGK^Itq%mx#_rl~u
zRRUR$-}&aDX0K=&8IEftf>K_K1eg)jyzG{?Gvnf>B|fe-r$q^oDT7OZIq;zxCUdQH
zwqd83UAeLb`6oUp`%|+*#vkmZ*>_@^Kda4jJ)!F?CB+0iZ-uW?Rm^a#3#we~z~2m9
zCvx6I_Utu{u(hZOc$hUpq@y5X3-IbaPx0-lOApcH-#ek(pS-_+^Jz#b$`KH(RG~7#
z4xw>8#MAeym=Y^-eA6rTaVbk+d0(~oZEdJaGXiX`7XCgyvynV@-bv2>ePwY6M3AOE
zGRkHN6IOYro|Gw9?;f3Ufrn^e-M6?--J&8VqE0T!D<GsimN;kN={Hv$kt;|@NRYFV
zef%FoZL$*1X(!(wfGz0a()2x7*c$U8<h5sQ%W{?Oa$gOSQ^s_<Gu|f>(D!)g<NwEB
zGKL7zkzF4srFSypCO?U09OCU_#YmXpJo%_?w%(+>p0E4Nt!)8&c<o9#L-Ps*WH3^1
z%SaCKV&!KIA}Q2^eN`w9+hbz8V;q=3Fc6sg`?|rC6<9Wpe4og!>iLr-WAeTX3vhKX
zx%HF}`Q!fD7clEM&63q}44;Y8z-dw+D08f$2idf;^@u$AC>$_8zOyONst=(jPl{7s
z^<mti)}a$sBK=|Ztj4OFZ5Bd;Kw&Xn%g6cY+YIc_NiFV@6&dcR(Cvu3f=!Vav+vc1
zSafAAA%!2zupuuF(4MLm$RRgR9XIYD;p(jx7=wjuRQ}rVu)}AwrBV_`2Slppw1}q!
z^tVBiv0cVM-n~PTq^6YkXNjr+130%2y$*KeE9@jO)D6!weapPCvAeP{*McuY@|xrw
zvNLWU#^GfQUDptq1}Ct~i0~NOr&=eU_=TZ1s#!{E<#ejJLrGENBaVT#!%e1))}oH2
z+F)~%;q#y==vslScC*NRQ;G694uPSo-C1X-N6*qUC8VzfubH~FoIy&sqB(DoROVA2
zt*bE76H(~og(HXpg+G;&veY5Me_q=Zf3#B7I9i*>{Hf;_PbM5goeqJy8Muuk;sFze
zJ6|?lVqp%EJqeT1HGgE@78BW)g(r>tcxC+V^r7)ph=-w;hi^<t$+hnT3D3l@WNDRB
z?q}PYmakcRUn4zHU+Xi&5=xWKq-!G4EEN#_m*a6DxRn>c{e^hD?V3;T96R6kyzO?M
zZFgonTI#$A>l#FRN_R)?h=wSlcfJvkFOc3Eo%JFQbg0)~8|d8fosAL={7$MZ7tH~)
zaM(|kYCT!o$rp|I90%S1+)Iz3$LGGX#olw$io2e(Cldjn0TZw7_<T_?(ss|bQ5DJQ
zC{Bq_{{*09;T(Ca*z7lA0>+P~8ut}r7+ETSs}GgBjP9XYvdDEgul1p+Z&Ps!LdDA3
z`o+mn_bauHpk6zUpY<c!14_xza<g|?=BB+X>iI=$gJ<C`)`sABwm4_PF1v<(Q=R7y
zuRFbFdw)M~HgQ?;oM>!nyCZa<licz<ly9O-jULdQx?*7ORKrzvKj6Sq=iX>m7qf~W
z$a+gnuxs(K;*%6d7ZU|u&uUBwlLy|RVCLWrUaVLmg)cr%$zD*+kZMHR<4`yW+g&l{
zuPyp&4Y>u*6kmV;eM`@ca~L*7_Bi%V9kk9JO7Qz`PK$?#Mk!Go+%%@ar1bJb&ZV|w
z01C?vf<@a67nVugdRFUa=|f^G=X!DiqVvcNuVA55$-j5I-yzf!c<^}K(!pe}_qP3f
zzD*(Mv<$wgs9#0<i&;CpZW>AF8C%EL>@2itr%DVlHJ-|`e@o|#?Cf*F&U$|&K@BW7
z8DmaQ6>wmma{`L;bxmWVU<W7NIt0Yhla)1wgFiG?*}tIii&$L==;pLH=<}^->NC!z
zA;>M9qYa>j4)wWkLwO>Tx>PT6t{F>e_Ib~n9t|X>y&<EM7*?Su;v*LbG6#QmAl_LK
z(yu5Q<mbDonKD**1yBncy{%3FC;oxLJ+(^@!bo#S;k3h>^j#r{)4x<p<jdFM{uo@e
zEPS-i;Q*4D<m81gbgHU6CLJfVgDlc-&c~=sUwZIuYvr-1O(cV?$fwUBdp0}e<H0A?
z0<Rc8H~I?!X@Lbi999k-mxpNv4h=QLJ-C`n8htJUA6w7Hu``As*<6dWD-^zn51+o|
zVKN6D_!sL9C~q5cm|@-9mu+))czP3za8Hk;3Zr<5#9%vSj=`#)j!29e8CUl@e$rzO
zE1Ga2UDKE>N3I|`2yWq=TZ%vFm5Wf$$-`<D)etFJo4rYTbA+OfYP(Suhne_Gb4gW!
zuGJmussYnLN2G=-fjZilF5gex@U1fFC@*~GFDQ&T=ao8xtYZOHrkdAi)aQc_MAG++
z;O?E}aR^J(fmP&CO2eRTq6n3o9POupJFW6j8i((?|90>4(eWf5$KT7vNdmjrIBvpS
z31ojy@j(+=-65fc%0jE(Nvt?hL%#aCED+3sbZtSc9Rjt;GveNM%`*A(aCYRU_f7h-
zI|Y55#|Q(tU>7GBp{hiBzaP_OipmcV767j`mzgd7WzrN9hg>M;W)f}-$W_+U^GRkX
zG*;|QVJR^+q`bV2EbcxE^K6igC4Kg4h`eG};yP^l#Keq!FcaacRhr9G0cpuXzFLg%
z&K8^jOl3ZR7{G+Z3myg0Xdz%|YbK{$cc(1u&{;x?ScMg@_(=6zgY!(tcgwD32Ny)o
zKDH34-N+$7T<V=-ko<qTHzNA*-LN_6ir{q$-f4w&{93_uNF!D7dJiN<N6Tmjvy6-;
z@2MW6BA&i~zh;iylsm6lv%={=^-H{dLgDTm4`~iKcdXs0ApCj4r!b-~w;Z}DA{fX^
z*iBW`5-aB{F`6=@8~VdkyI<0GyDrgPD+EWAbuqbC^H6O1AmM^?h~z1gPDk{jhZKci
z*pI4`?prn2&POZax-{KQg1&6;r3~nTH=<gLCx^7o)m#*=H{K!1hmBvNF66j!FmYrh
z#(Oeyw{$~c-$38KSM0FbmAcSRnq?SJyO~zAn@HM74SA09rJ;N4QbE$Yu(*}*a%1st
z``Z?bm-*diJV+mBs6RT(-7`qaXLe_zRLk0hvlILC8Nn0bGxS$wS=CRA^b0alt)lvq
zzTK2g<9iLP^HQ6t3S|@)jK2^0$Hk5d=hXOqSY9yiAE$UCUf?8<DDLQbtXi1q)R0{2
zBT}@OagB+xB#N~-nk8GQ8=v1|kjliiqt>Y9G&!pRWpq#P=F$Kr{K1B`c>fYCA;sj=
z5=^ELm#j$D+CKFyoufqvX9$tQ2EFQrp$Wwk2VGb5=pnC12(qj&SB9Gm%vzhZzEpjq
zB!!y(TX~3(VNG*h-Z#H*Y?`v-Bv3EEJ6D-M5*c&Nj$FJFK;OrZfuGcgN=<)Ja=mAz
z`^hJuC1R)COosg<aKv2tuqITdGMMZL<1_Xi{Yc%;&>@(FlGETq@ZLF#ux)yW=c*cd
z;fH96S{9Tua+}Chych{?+zqTWfBJaKj)H2fVtIGN`cMvSMV}fCAERt8Pyy!8<lGAu
z8@CH#Lb>%U3H;b@jFvY-F}yat3<}9{M&$VYX`W|~hn62UhH$f4y?;s#0YclR_|24g
ztWBq9%MJpsZVm?34nb!+CImzF2g2h|A8bdu21y%N2On0uPd%dOR&)m5vD>R$p!Z6S
zimzUR2`D~rV%rl-s@&|22hSUOWfW{u=Q1rIGeg9PxKK`hM(B>CiZQb32sjA>?{R%j
z=AsZX9J42lvxeapbB*PD`ykh?zaKttpLj_TV8YOJCt7Q@5MOX=vZ4!eGRE%aTRAV~
zF7<N|;B@lJqGDjp`80L6Iu1MJvWa5ZK~j_LNaZ>|A5M($`(1VlTNZ+6Lc!)_F4G);
zzOBv!N1XV2{EFILs8=@CN6Bg&oO9!k=fqfdG)*_IN-KGD^T9=Fzl${P)1U8hE#MN|
zK+oa`4erA`KIpVvzJf9=?9$ClBX@<<oG(<$!ak6~(o#~^7(l^dsBf1KYMCI@p}JvY
zMv#bS<p=s;?mIY<S(~C~@HS}e^^6ZIbR)Um0D{3p5h~~zvQjqrQfylXf~)40jU~B^
zhn8Ay2cyz8<KkJO+Rp?q=u8@)pyE|8&J@3}M3%VtJAX!wA0CeM1_eMOG;2`6au<S`
zCSHNCsSu*ipel*3Xt0gywUI^XXAjkdz4?2QdTUizQuZ)zKj#q+10Ub(a$EEf%%T-Q
z#W`pzUx}YP%1ws|t-IILD-}LmqvyrqRz0Dz9qs1gH&1=7hF5^{=+rCAlb9KaKi|I{
zp*;oJ>~XFd&)O8z7&AhVpl62U?Bn?&qRbjG=N178?ToF(rtvUUHj@&#IeV$7Uxh;k
z0(dfZx*G}>Shlrv+BXk#4h?cyAymde<>r<SBZrMA_apHK9Q_Ol81k(O^}smxP^@#w
zTv_d}lt=Ao%utr{f9fQ<2oRd{v0{+t>5>>k*UZKEB549&g})F5UWGKp@GT+Six3!|
zBV=3=PqR6g>Jtji{)FKrCR@te*dK1X4ugC7^-%(<Jfk6rY>9}1Ca~yrfaZ1En7@S(
zMMK$kPbR+^L6Jgx&zcdW38DLw=u6nPyzO0l#$BY&&zFL^p!dkM-@lQ-D`XTAThxHt
z=3a4JfFbyMZ{mjeqr?v9ycWgkEVC&QyZ19bC%%jq%25WI64~$v`JsgeIGuuhRw%)&
zlHb<oGz{ChUnY9|Do%2;eQdWY>>k_$R_>J0w;lltr$qZkmPjS;{dE>foPKRkv|VzX
z`NE)TG@_ZzMNjr|j}-s@Vo^<?=5SsH3Vwmm*ESFFFn|O1qH&Y>XH`K1ngswO-EAn$
z^i5)B1pu#g$fE0NbiOzcEH<vcgO7{rak9v~Odj<I&yRS(70A$dy9c7mwN9uOXskfx
zM8eMaqf49^*+7IV?eYsMeow(MICC4B6~cq+<-84hvJB(lPY%3-)39iwhMHlLjeGv_
zxr2tssG2RLk^H4rDn1sLs%1j9#G$r53cWRNYE^fL4NH9T`S3F(nHSy_MAbnZol>Io
z&WZjSduYlfF%oTCr5~T+@p*FjMk=NFr+b3NzOR$SUTxzPG}PZZ{&-i`z~97>CjP5w
zD_APT{<r5FeweIeY=r(K84%+M(-8sSC=~_bwcxDO*{cF1ul@$zulP64q9m&r;Zc{t
z&#)or6h9jv;04&#ya#?lSx+BsCY|kSyC2Y&^cu;wO-jc<#0pB^U3y&HZN+}Rpt8CC
zRat|CloxN_BV<MHBUA52u+CebPKkI^Kh`jr!$&qlTvzou0=>map&O|~a)U1hrIV;+
z{--RT>{O!C14}a6U#tWa$tpLwEUsm4@hxnGGy_j|!B_jt&vdw`UJ}{#zv9NYdosDv
zM6pJ1<?e0*Bo?h<vt{7IO*^H0y>JfTE?>=%AYOVh*X+jB0)#$AJ9vyVBfvIaD`p(P
zc2*Nw(8H52z3(Iw@yEE<_^dMs!?S!z0#1ZFRl2qcj9rPU--WS#v&!g;Z?li#{eNtI
zWmHsc-!C!9fW*)sAVW7uHw@j4($Wpm(%lM3Nq4t&cO#ujNOy<QaklsS+|PN}S?3dL
zvG~BuwXc8tYLm%gAnS;W!Y~fB>k=Js)4Oo};uG<2x~+@}Sbm%o$=+2KH5n^E<zFN^
z7f$I-OPt(bRJsdUkcdLjlMohWeTMr*`wyThh8YRjgAux(Qa$#!9JxrW{W!ZWGI-It
zZ2=_JYA}}1&-%fEV0u!jmEXO<YNh@1YV*raCL0=rsD(tSAKTy$J<hjx^09h1z?$&h
z>vTeOC?558_K;dagF?Og?ZeIi+WHRKPs^^fKZPr7z4#wis}^)AVxi%Eil%?DC#Tm4
zcu9#u_8Fyzew)-0TuQ>JMO0^RB>bNZ9yUp)cfKrH-}%&kEP6ANY{X`!(=rOJ4QSt?
zDR4l@&uK!yi1@KfYo>kiNpAO4t}t)i0-4r`Lnlonk#`D&TC4DmHUoIIpQT&5$o$@i
z`tf4EzL9JN7Ku=@cb04Ts()IB>UuWVdGjMVd6Mh4BCp8xZ(zJn=P5&xpW6BKUxX=t
z!Rsb3l?#iGJma=5fB6gA^LpPTohUgt-uk4rMHnF@G?bq>0IK(X_2(!s_H{Pf_=`8a
z#CaH7gDhw86b%>S2$~Zd%kDQz-G2#V(MoTZPQAS5XJ$XeHJ_QsK6~H9PM|DuAeyKZ
z-q~3(569qO0MG-^Cg{UtWfMh7{pn@_jQLDIGZb3W;e&nz7fsoUHxk0$J#z1zdt7|%
zFd)G#7)NyJW}bEf4~xNMmPvptSYQ`}mbUe%xmr)n`b&7ru7SRScD85H#M7{bie_Y6
zhzb6k`QK7bmI&pagCwdqJM*J@q_~37%PS!F2J6u`aS1OL3OnY88-kQ=K>-+Tc6LlP
z=7?shpP=qIJ)}zmAs;C!*t+uLqV(#6x8SefZ2hT;YQe<$Esg<SVkpi&p82p;B3Jvc
z#wuCLXgmyOpi~6GpfT2Le!PSqr)EGgBv<P=-CilqGNfx3yS@2`sS{(~j>KsI*bM92
z6!6e2jP*|Yaw9DkCqu8-dt4(%5G~Kjga#pe2%cM#UXCz~!@fu1qcg-bG&fkM(c!?_
zyQJoC%P~lX%D&mUR5s9@z6vfc8i-p%=@wdH@jh@>O{O3Sm|~H=!n?OK;Hmg5<K<3u
z+ugcb4s)9rTWYXlWMC*QEj9kn<0zWNtVW(6IwcDxN%fBj^fU5(oANsss@mYCb%5AA
z42{w(ya4}>{0e~_oj^<%Xg)|KDD)a=8^ar~-YGfY@_GPJT7$4^xu<QNOl+UqxlRLX
zh}i6gP(G{P>L1U=d}w2IauK3V<}1e6`*!KC3`5)P?7)FIZwQChKOBOe2BklDX>{iv
zMYdsUK+CfG-t`ui)kDAxRS5+<_?s==4ybRc;46ExY+M_vb=N>g{bqe7z!IZgo+292
za0x<ErOlTtIC}Fm{cxi&(qpR#8;aDI>8`I4eQ6kBE<ZwDVEN)83bkPGF=PP)-L?>?
z;C(g95RLbCF^<=6%viI}#&#d1SSW%#hQoCZtWnd=KlkT3wlHFPtbQLKb2{ziz9C}j
z2^HwmiP9@w%ZB{aK5;VaPcTxkCuwfguo`CX*DxJPCb#pL`u+iagfRRZ$^`EUaW|RI
zQLb%@ql3Z9wqTG1JrJytGY>^5P@>#Dxh9)#3Z_u~H55+zT|-e6s{6^uSOI;BGaTm6
zlFputoODvfLr$@%W=fg60~~gE;pSHwHiEOi=bRg}6;|G^2f%)37huhu@g>O@M8q6S
z?qPO(32Yp1rWpAVS&2Jfii3PquV=TK!(szQFo<lGHqdW>YYtR6rnqonsuxmgr3(qr
zp@eyP)NEyT2MTAlwAYXjSb;yO?FtJ-E9jkX&;DYZ_pn>6lV!rEjQ*dh#ejgmY?}S+
zXM+}Ph_fX^pcvHxBKNEJDHJ}Q={M`KJ@;YfWg^dYaaRr_c|*T}Hd$))Qbi<F`2o(#
z^4CZCntOXrm(ty>-M6nK+9R|RrOQQ>ADYw@n_pf;3}+LHW9<Q~Wd|>BUmn`ZUilbr
zpZ<J~@Ed=41OuubT_vM^-FYc;lI>_fcZi=ky;e!{Vgtik5>(2nceyWs^JhO%r_Yfs
zrIramX922kbAByakWB}4Vu#vNmq3V6Py3fSIe^BL6Pce=Im2bI0gXb9we|NaP)z}?
za;Z_;$-9PXXJdPG^Rgfq3~X8(iw&`)&0;S3U>qxi;Bo0eGAgC$WkBU^OrkyLloV6x
zD0%1(>S^Y^>8|7lQ&CnZ$LbF`P-Z85c;n_m>}D=Pdl!4kSbWf%flj-CJ@^A|B5XMV
zYV!#_z*>z22cCny(xHMB;r4vNFm&A%j03|UqY}C7t(-1&#&fLqNc}J-noU{6(!n7e
z)u&8AhxemBO<R4|FE8^cvW1Tyk*#mmzprTEQ|2p59h}UoIc%;TgkUUyV^cVz^$&ii
zZg~J0G%$^D>NjtrT#(Usp&*>sBrG?B;zi243x?rI>r<*?F>~a0uYJlM+m)`@!0Vsm
zsf^V{Suuq^{^Fv8??!%q<%aeN*}Z9~jQ?HIb>}%KKsde5B84lV&+Gi=eDj~RR7o=1
zN&tIb-B`+S`uk(P<BKdiMHn<aiB|pt8Q~W{3kR&Hv1*hvSmuJLY_nuR*;UarQT*V2
znw|P5QBxPMZ$B}27^a;*=V>&vDwaC3*Hz-!@HE4w?W1^aoh#-;kw3tj^D{0kp4ExT
zFW}7IZmhi#`quZYFm^XX`yxv-?n#xqxA&7LZ}Bj3+rrsTI$uvX!PqmHt{?P~VX;6T
zH2{j+nJF=)XQw>WGZHes>fb4U9e0a%<MTuEU-+~hTvww+=PlWaJ0aPwh<djw>fV)@
z*&YsuINiIt5q%e*P$ruFs-Z+O`~4U!dnQNk=b<Ubn(s7L^bTKrSY%nmVfXMeR3(-z
z%q3ntbEpzz&oThv_(Faxg1ov}vlFjKjIh8vX#NMQ*j;EszgyXXI<36@)HI(sZPkvd
znhjK%w2}Oi4q=_cOm*^Q_J<MleBzbZiurdLse+1)J5UvX$g-bMB|pCmVUVn0q?ojL
zIov1QMJR#jZg!=Y*J=kxB=j)Kd+t?>XTYL5WuS&-h!UAkxmp;s4=_dD^A(Sq-C_nT
z?}P-!+{I6M+XlEvM`8Jaf>~NWjt)7st!-|^a09PB8&jM-)N+=h|B_6W5`QjCtym{?
zK3cmjK=y(nysYT<x3|B>$2&iaA^}`aB~uZ_WTH6HH3s!_RlA^zeUmI{sz1%mB?&^E
z%$u!Evix0Z=HnQtUA*MBj0rdq?v|g-!-eVuzb|Gf-;;IDdAHW*pkRUywys~%-k0N8
zEm{t~4iMIt@!b10FY-P`kC166&BxR0p!TITmSu60XuOU@tBk6ANlG$~E20k|Vt}cT
zjZ!PB4XxTE(tmJWMQeVA(A;!JL3k`bfV56rFCVAqAGPY7@oz<ciyW{u#-ZrvTe!1_
zSushmt8~0=EV9|^yC3GXiy0#!-B7X{mdd=8xm&J6PHT#N#}Kg}D+2+8Revm<!#Zq9
zL8w$W<hT8c<i}D7f)-ctf>s>hwA&TAyEJml(w1i~HPd?`y=khG9`ci)9}(b+1*?io
zIW(E;8)z0NWAyE+Z5r*8>@#rL_m|}=)%vz8pen(KK@x8bPHaKLJ44*)1)RxRP2(5=
zFkUXM5^TrmRWRM%a*qMy`2+~Tjwsa(PSf4L-Usw)X1kL!LPAF0d?-I6-}7{wyxa*O
znStdatj=RcVMJ4Y*5!SnrQ&P$QLsXOCw!1|?rbvvvH*K>9ix8MA5Wm_0}_bbxm{&O
zt_1v^ugwXJPw$I3%(bR3<(3iLY2zEIOVvN!xc4Cey9u;*!Ta|X-+#_q1t`BJc|*@P
zLA}EpX*Y&82!S2nQVw8-ql2AJY9FgzPgnzG?|xG}*gsbVn!*(qt62*OjFqLlGy>qK
zHVXDm4p*^Z2DXuDtD3oCl@*lma3!12u(zeIJvOk}pn2F;>$d6lrsnPkJ@8q&fySp@
zW}hK%L4ybjj0@>fCkyQ)>_zHv42UVfBaN>rVnpfR5a*5L$z1)=VFm!i5pEVRuO}Ym
zw@Uqg@AZ#B812XNF2u^8_Y_tl3*uoN51&CN(v4W3U=3TtielM$Ps<%ivfNo|!*qE;
zr&z{0wJqJLAut(w9cgokvyRv*_%50|M3ysE_E}`TSmDnu*oZVlp;&tGx&DC|ZSog0
z_{2-quSBzF?uX;P{R~-@`bZrm57$6=r~z7IiNPd!AvB<jj2s`(*CW#nlo0)S%221A
zZ!x$|eSihG;d2Ui#%PDXewCeD{8cUkN@rQ~d^E(;nR7M&@{%Et(eUdo4Ft~JGo4o<
zR(=??EEWb&Qb(Zzj8c6>&nIM^>oESP*9r6?8HF)#UIt*Kmt!dB>0cF{D&$_SSE9-n
z&;uN*B@_GTh>XC%QPOS`+7pB;9EKXQ;c#p9pCg6*ZXt%pUNjo*Kmot18g@wggLS@%
zI+ky@!@{wJYyQ55MJoa=qr8^9^NlJ3f39Tq6EWJAGx_h;A?2vEO}UMHXCYb_7SL9A
z@9yr`#aQ71pz04#b3Seca3;pFhc!bWrqA_C$kt*n1pB0JQ{l|7F=XA%({KFwF{KB^
zQPK=p-zqOlnak}7hcP(^Z=~gN*ls?h{pADY|2;&YIDgv29piu~c|ffK;e4SiMHLVL
ztqVMD>ea1KZQ>#ph1&QlLBQT>KO)9T`}8hw5s26DRf3;Mmjm2n{J8mfIJ@J(aTWLh
z!fVmU=VtK-)bCJHW{q(AyUBwUo+KBQyF_;LnyrKHQBTCsh^}xRKztPJYpge-cLHvL
zKT+*JwSc!q$UeVf->U$Hw9~x=*Ol+0zH>IR*ZN}S3Z95FxbfJfSlyWrazsd}M5h;A
zn#A>5tiJn}6;62&dV%wPh7#X?i^Cw2G6UOD<p}k3`KM$@ZapOkA%2ofZqSaaU<<c3
ztU$>4v4WsGxWpWV^6v1cmooT#p6GZbjZ&KP^2pt>O*rc*J1yDF!>$y!pu<230?^_d
z2rv66{OMj%P2rk?E8t%*-eXjQ5$J1<*{vY~5)`WY)#`s6wD!1E_Un1s^)*!URS*az
z@>V&D8qWXuuJr}2AfJ@??*Cv2eLG@{atcLKYa?0$g9ui(&g5MvM8hVoPCV(6+ZmJl
zYkfNi2V0GEi@UUPCxpokMXqnw0Tq>VP`njSX(iWXzp82A*?QQv>HN{NR#I$^7r~Lz
zI163G%&uRJe>xUngi@ma#89Yu?|Lda3^?1=f2XlZqyWS4kqNRbrBH|cTE0!Xi|cR^
zH<r>wKSSw8$6@v1mGlCTi0%F#^qhy1_F%P`W>Xw1ZqbU-&0+I%no5d*Evw0;*idXy
zQHvhgc$p1}8b&uhBVyj12k~QLCO+dUv2o5s2aOa=z6=J&=Psqw-D&f<nvxlJxZ1{-
zVKTm+dK#y3{;gD0k~IWa_C(H2Ch&c_1A1}rt}rtd)E~(C(g(GX?v`)*+1Nsrx4u{^
zvMwk@%gh)B+TT2d6;9PFRrV}mA^cf_rMjrJL+G_@Edx7KJ|Zo+zsXRV3gz$Iq3JI}
zSPx+;c>~nl9tmt!pXR2IFOGdKNIs$C26X63@sBTHf^zAVf53gIC|U*x-_%0^^lQ91
zC<01I&$J0G5g?zcNsodZQp`N*ls)$ABdXI}I;*MUB_&9fIw9NB$j?NcbD%^a#W$q*
z8B`EeO~I*sn%lkB{w7`mpl4S&A{c)t$|e<-%?sPd*Qk09KoNZrC2+0Q8G9FAUf8rx
z=i2UcQP;F1$mol<;F@t$MZeEEW{#HkL<HEiOuQ#q^QpbeuSI!D>?qmIwB|#LrnYUW
z2>KBvwPm$6`98lA&u*apm*ljOfTod?AS<oe*N-}CeoZb5?>t($lhRkkfFJPxT%}FF
zF;yUHA4SsBFtU@=0k7|XYnm4&!re;~gJO1gnrl8{m`6kS$GmKq+UA9B=G|Mnd7@ZK
zotfh=S}h?i5>z&vLWUU;IVX6QiNO6i5Y}`QKh>J<$)jZ!^eN9+wgnWbOE@)|4VABc
zC%F<%%r}kFR-|Q<+M%P2q9}P9^1{<L`u8Cov}qP$o8EcSf8pV^Ni{~yR79R3BUbdC
zB;7D;8mKDo?}vn!M15ZXgDPdlS(IvY9?ir>ajlw7H9cQ1pfGh?@E^hg3YhwJKGlRN
zmhjczmaI|zxlr1hy8I-!9(*FUL($4*7&FsNj`wM@2K^*=&Ohf#bLx4lkUIbdsDRe3
z9v&oQo7jJx>$tr67X+T8h!GepNv>y;PwhkLpeB9rF1M*d^@EP7V0fgO9U}z*7cvI_
z^`Rd4R5xPpE4`YdBI0<i&_k)Hi{TenEOIdc_b<~@uY*amn4lc=0eQI^&8>Y46Iws!
zh-wyMk|kTF70-BPsDIkg<JhR}aV~GY`71G1u+tcqf))<U(LnY_r~$1b&Rz-!^B6+5
z*}a+=m8p}6Zz~t@ZkSOtg*JSJ`hNiwT=8sfyBd+<Jt#kxo+(e|=$EoMfgP*FdV;^?
zz*8`n?VU}=c#IJ|*3^?#_=R1@@?NPZ(6E$4&l7Y6#?dKAcR%bCT%RLsYl)~|H>2{a
zAFHsM7{7Cjp}XIA?4yMN10Q_llWuLtdf}c6s$jd&pSq!MC#fj`vMhKhfI81Q4`6Xx
ztZ~4mGL1vnFSlhw6zvh@Mr+oP&S$l%QD$d7L3P{|$Mq%CM!J!j)u!q}|7Cq7v*$>P
z%;yKnXOrYpdx)f+n;DzqbM(HBvWYu0w>jU#vSpWUVnymg1{>BU2t|DHG{PHR23UZk
zrgtv)#cI|jX6C7bfcKrFQ<#Lq%lH#jH^Y5<Yp0{SyI(`>882Jctz=j_fzNOX$P_2m
zg7;F*)CS|<K`;2>!R<WK>uO0M<PddYJn`)LuZNT|LmFrEE(@19)Hn}{zkd{0a@>yn
ziaYWpMq^b@Y+I+%ZKF*w9>`XtLc)B0MhhRLB*h#6<LwI*R1O`tW*&(xxLJ%?yr5Yy
zkxe4%+m-N}Y11Dj8&luFgOpH^LfTE(JG2W2Xs&!4RJMuJtX2y-X?gd{s2|?n9CdbQ
zD^Xt-ZEAgMj?MIP|FxFkG!YZrarh@>%%<B~>)+N3{H%NXy$O}vjV+)oK_|`ez|67|
zIAteH`A08OB5FYLyNOpX9y8|`^g7L*beh0aPU&1r1V_#MNI~IyP4?3_1#*r~A5H_T
zlJEL$Jlo$~7dwCKzY8f?ca^l599R`!Y@6yRTzzLlW6v{qaFKMR*sc1)p8My#h~gM?
zq1n2}JHt5iM!mlRlf`17UsURm1<aGl6n&m(L88Lfk_+lzaA8Rg0J-MhpiwpDxE%ov
z2_8jzE&e$wDVt^pcKnnR*vB*ifSneD5%!JdQWV<nykvTADrk87$O9!7>O#oYKq5TX
zQC`leOk73o319Ysjku*82W=lS3i5$3L6jcsCCcl9;c|eZB%E9{%E-|Fvgh5WMY^m+
zEAYS16o8(Mt}MPpSjUr>M$s)l%~{S>14OS6;%PQr94_I`=DsL|R$n-GwHaSOY>F02
z4Y5G`jhaIWuG$$?7=H0<r6lG`ck9NmW8*Q~jPw?@_yYogdR3q>=r;=^pf4Hc^Hwuw
zLdHkCs9GR6ggX=UE$-sNua6qFc7@2=7W^Ci844WgzD3{@t=r+Iy~K`00!R$Fc!2Zz
zNj`)cYYwan<u||2z($F#I-NuTvd;~7pn>0dd5Ukg4rf05^Fv9_UMM=QW}));>?Fz0
z`lmZc0^*F}9ZbgApRJmSGbXVIfI#@iay}e`w-s>JZa!?L0+<SJ*~DO3zU1BJ|E*w?
zucBw(9PZ7(hZHyK=y7oMM5b~nFJ=AjkMl=M;hjmVUvXEuq-Pw~TJ^@5+x)&zH(sXd
zz6+l15kh+Q5QG(ROuV%VeZ>~E4*k-nr=p=}PEVg!_X|x50m&|UylV%LgYf8*AKY(S
zcVU|0rH@4U$T-K$FE%naR{c&Zu%6v&fIsJz&Pqm;GX~~M#Fn-B1gmaj4f?^BI~MDg
zGUB23i6vIo!TA3Y8@S?7^Q(DPEX$wxo4<mFZ49#_6{03t{^<gxfDl~nYomBfy=JO<
zXBeQ%a;)MNfRZyxB`NEsEtLUJ<ZB|?z$ai_b|EixvtWeLbYTikSmCq-jJagCXqpTo
zL~SKa{K7v~&1ooU*Og~_2eDWD;x@u`8N;fp`eUMQvq*h${!xZ@+k}b-hTRH5D(apt
zV%Oy|sAD!G2wMe*%6unDqWEG@)Sis+c~(nxf~y56&uge=CETwSiXu4&CT7)~$^Xt&
z|7gKnAu@n~URPQ&8!xBCuVN14XR-YtTG!u}R{9nK4h~BGvtLaS`~z?d_|__9B<k+c
z39}KgDkKOBs`1*pvD;}`wyd?yc@5KWdlYC8NN0o1Sh^5bgI6_eoGau9r)eG=3rFl4
zZKt^}D;{NufhI&R*>A>s{ft3*x$XT%Q&cmpAPs$S0C69t*8(H)F5Er;y80?f7Nw{I
z-qs-+DRCOO^R3A79S&5EFu>Ui%3{_iQA4$&oCZ$bPJ1ykW-f3Q`9A{%eJE`7s3MO7
z@?H}xM<Vs7j^m&5Z;b~^ZJ<1Vf7ld#9}3PSq`aHcH~H*Sl6Wk?FDQO>VH#neXFlWC
zGw5tyoaC4#*yu_et+LDJOY`c{K$RsL)`XS)(i;c=NNFIdYAnQ%l49?rjWQqA1{U%I
z7-Blrj1EY0cXiLdew4ZQZRS_XVpU|SYCH<WUeW^#{jjw1wH(Po^IWE1!a*9~FEEu2
z`K=h2sar0?K(cM(?S;S39yZ$eJ4)QQ1vFXo@@u}C6->}B!kYfyoAQq^o)j;DKk%Qc
z!G{@eCF2Z17x--&pZpRrL>pj|CJpw&KCpcNp7zer-qYU#NZNqZsvJdGeuP<6<9O06
z`or{aAM^mncO|95-7?r=?ny(@x<b68oN3U=QQ`h~ACr6)ovw{$)UP;O4dHjI4m^S_
zKe}lZN%rkPvyXLZICP^8r%A4S&AbTcf7mA@&~s-^H74TPUBkn7JO(^bfyZI88QnHY
ztjXs$svy4_$aargeFciovJ$VbE2c#lXWN=75Te~M?%F{7D;rMi&gJC?e%EP>_UN4=
zX8DYwk+h{xKb!s?@c{^jd2m13)des<`2Nqe>c^s^Z*{*L?*pz3pn&mo&xZ(}jj_AJ
z#PFk%v8DiWGuRT(k?LVpMBxn=wA$Bn*{5=2O10<9E~9#d$og}Y*d<^H_#r>%pZG1Q
zSanxg3ROfqJHPA`0Az;c3G#|LsFShxJaw-JS~6>^(MZHS_8a4kq1g|8vHlr&bee>d
zMaHV4v}}IDDH$lp5fleu`L>*O_nJXOPKNP$$-086`lIY5kpPZ?b9@i+6~)Bi2X@fO
ztz^L(*`o88NX>oH&z;ZidwXjs1KS&5)3lcX4F$XqEHw$PppSo5!N+0^QaM4doCsnr
zezE^ioDW%B?j0J+0`Bem42J&gp0CK2Xe*-=84j?q-LSQTl<KYTYjIM=7m-AXu`4vo
zGzk~c!-Ng}H|v7^x-(|NY!9G?+Q1v_szDv;Cs0HUz35|=*e_mDF3o>$J<uW-bdklO
zW=`uYY`5CDf9&Nt6SVZ7l#K8%EZ*<n`Szn#(ET&;F-OBXzfDd5p*tm}fOUB~uk!X4
zPDJ7i5)ZK(ca~2B>S+b>twE`X_o-Cp5w1rE0OK$XPb*R}YpNwMBqWr`N!cbOAMh{{
zog~Ps2c8Ys1;&Q(8!TrLAti-YvGr`YWT55FHz|;iR{sI|F{@G>6eRtm6kpKrL6z6>
zrp+RJEt4DtmT|pHcQ%yTz;RwZC8mM-aJ5CaaoO;64(PH~E8(8Ajl)N(3Al3lke&fi
z;>2o3WO(`oX6gBZ1d3WJJ66hP%9vg~Fop!E0<5_0Rj#dxF7{peQWldl-``b4w-{k{
zVDI96s)D(MyJ6GXZG^D{YbokeUs+E1(b?{2_pB1v!5|Nb!p+5%NdXML^T!<Eg+?7R
z5oxW4ZWqUh!!2b^q)j7l(v%m8zWmOs<wniHjc($ms3^FTV*ih7<9{wOS^$19Ri|Le
z0~6{8N3i27jZXBfT_X@aAZ*Xl(B$j|r2f+qNv+U-CKw(306_jxXwwYE7bzZ|$;Tf<
zQ)sbKLFicpDpT-Op~6Va*o(WDkNI}y5zWxX7z5?eLaz0aIwPaNP9k`ha4O=QH{4jW
zSK8m?-)!9vTKuvheUESeWBMS|7<t7twJkvVfrBv))e%?X7+emB4@B~*=LiqX`b1-E
zTntc$OaT*%C#)ZkNP-3scu}y;f<hp&Gq|pDSQPI5ID+cF_p6{l#w;e4fy2vnXGM3N
z>*h~nnx&dyMeE@=$bQyn!*EZ)I&ZNP&*H6_z(^Tmx1|gKii;Ve0pZZ4haXGjhq)AH
zGu<w4P*|9-JKvY~JR<o@lsbu);+nt&M9Nh@E6z8URa`QvWMM=F(1kN9&}IiQu8q-=
z;jjqEtuMBO#p=7EZS95!O^|68i{~%Yk)q7qdwhLiAlP#4&;PB^V_&ELKLp)&D%gb~
zkRxgr!>8xiCwA$PiQtNc_92DwY>R;^dLi=TLaxXbT3#~;vNcD-ZG5}w59#G<eQ)C(
z3YHpT*@UYtrI1T{EY59AB@W=-0b;^s$W9g*aqU}ln@z>nO)<N&o=3ZSizFxBf~Qfh
zDl_+JKdg>WiO&k8`{EkRn-}Qvz4c?cuw&Wc6aG@e>Sl_pSyCQfF;@{pJ_#@~EF7lc
zs*oB^YFu}V29j@QnuRh2JE2(N=0FXNhJ-0Q5&cgS(*}t_<(K^6UUL#kN<1ClP9sdZ
z77FsEH61jHD*NRSq}@V)ispHvR6-FcbR;GKp+BX|zz*Q-a5_}EL69dhcFb?%|7t(C
z-`NfGk;93YFuQjT7xO<MNF#y{>9o-z!Lz9W5PA)&+|G$yVYd*WY|1F+x6|X2XSZkf
z8*Re4Hs;KPS53pPG+sD9q(rqv?mue|<fZw51t8sSsEW#5o9=kcWpCC3C<~LSY>Kgr
zln^DAM2&Fy)P%7Kv!KH-ih|#5c`%AtYsb}KD)Xs<m~eRcT?#}GqwY;#vL$p#>^T!^
z+Hnu=>rg8F3unJF6wnDZCNk`^=*;iT_*&v~+j*s35bNRm_f!MHpe=~^h+XHG%jS@!
z3s$v8W*(=GM#pYt?040wkQkbi5&i}54?7R1JI0+`(Wl#&xONmn0s5_3>)#tLBV^T^
zGcaQ%9H!2N<`d|>zmH3Yi=Pq?TgWP(xZt`mC4au}Np<JgmHgtwxG;;cIX-ZP5nqJG
zFNWT@3}bg+?Ymv=@Av7?0B>_-zk+?PaEsBOMG-dSI6^<Mgk%j$j#6XEr07YsmQ7*M
za!b-V#}ZZy*G5rC&FmuU+~@FZ)w#Oo?b8ufP!~LGasbQBdGv4{@F>rujCBL+eiB6=
z0J*cZw6)iy3r&+OVMfI;Blf)X@nn`?+nCJb`6+<mHGDv!coOM0XAE|kTO=QSb!>At
zrhh-fF!nnmB3$!ef+YNOy^b`Zp@U~^?qV{*Q5S$8`T<;M=TzzA&55f+JrsddWmi?^
z#+F<pG^})xVmk5SJ1#wim91}!zpXI0oauPoBpSxuhvrllLt=`_0)q82!`?L4*!{XP
zyoX6dA5LZ2B6F!dKXtN=`1z}edVYT1r;eP!FzGl)0nBK~<i`k+AJ*_FU2jsmDcOJF
zleyJ=oDU@DV0Jx)O#6n|x+iSw$^mwyF4Y)>M!-k<*r9WX6K}+L@;XsY)ef+L&4jJB
zx=ZuYApbiH%~4G9N6)kd7fkF1W<iS~Q(s4PN4}n}a|Sj0KVx_C-oR_V7Q4`cI0TY4
zRW!!iBCpV*OCY`08?{JZJx)LElnojkc8II6bk+ye+$lTNibK1w)vU9@<w&$4#Lach
z%hkX1zNW@M_1~^q3CX!t8Vbp-UJ4rJXVbceHGS|hpg}q(@N$%xVh!$7K_7#h=rIlF
zz3Q@SO}iPaG$yO;OOn4yq@|D~J5k^?VAHHRW695aqbZgRSd<KaMm{2}l3<qBX{qaN
zo<zh>HI5-C7F;bKhtBTtvF!%SM(eB-WrO_v3zHXQ9Y6Wo#=r`9-}%*U@3&uNC65CA
zFXrcxx`hlaA3Q_faFY(Lcsp8QDxu|DErv&J0yf1Yq6=E7&#y_TkE*rs3U+oI>&*yH
z?F*;ySVpRt2a;&y_a;q2THzn;<lpukTwuKw;Sp8178ZK<dmDQWUNJKs7EV6-mX5EL
z&>$aB<N$?W_e5^sdiQgG>5pHg@27~au=1Ho&PMu6`@PgspGI1OJS(RH({XB<00r?k
zrPxySpHancUs6CaiR%Vm`OiFEzg+Yi4TYH^upMgX*QpkFoE3{cH5SGyefSAD^RS0N
z6g^RSQE)44L@t@z?=wxAV*2#LpB-&C?!{?;40J9r`wYLyB%qNybE)BJB#i(zV6rjI
z71KW@irYS7?(e!#l-B!YqYD~zcU8mZcYJC!e69_ByD%Su`rZW>s=YZh2zq<J^^lsQ
z2<&bDMI7k~B(k@xmdC?nzls@9^SX%29QFD3h!s&(2o3($yK?&QV<uSPs<T_W!Qm}~
z@t_!(D?qtnRaypdf#oyo-SZzT@E*b*!mVDM<J(u2F2s%BaIB0MADPxEND%hFcLZB>
z2_A?pFK8(}VK(AL{`S+GSF@q#I&VJ3i^IN3tiR64jDM8$N2aefaF8T!{%jm5H7Gtz
zlp!SXY4THZU^5j0u&}!Kyh0rNnPRya2CPu$ACb8wik+@UoL4fbiyp-xsmlVvk;?5R
zh7W+V$K|(mC%UICMK7=x{dKa5h}P^I@`rDqPxT!Wim1zdbZla8TpIOMdeN`^jc@O!
z!hDhInfp*@jk6aJ_p9J6?wE{JDN7|{4szXofA#sn0`ByzP}p?&GBG*wAy@4PN%)tr
zKM98E#Tl5rgJ?xKy-1~iciR}4X(}O4PxVSaOk%(DaO1?Jrx?V=Y6f~m(czP`_M0=9
zB6j`sWYiCVct0QKhzp{J_cXp_V+=I(H3-vuNijLk`#Jz9LE{wxKY_}&D0DPJ&G1kg
zCR2N!-L*4OM)Ct87Z6`Aw8T$c@$(Y|Mulh7!<)Cnjb8Ya+$O3C_MOpn?Nl+p`1Jk;
zKOi-ekonWp4!qM5x@#8!iiM)LQqn@>JJskr?^rp*Avd}ZRy3KCru@-#tA%evM8`JX
zCaJi}cw2t<(SeymVVC7vTpEV=W<~S>Jk#*M*k3djBfhe7cavi${4)b~Icbhy#;Y9$
zQy^C6ueL8vu%@$lX3E>RKJoYaB~Jk>*6FCVaoSnH7A34p3R={RH<rv^D=N>|a88)j
z)?1^bdNg#aB&jl06iiYPXRQXsW4P?zcara(JY$J(`#dYEeH0w}dKXV|r@DM~FZ*fr
zcx#d^%efE*^pgQzglo_V%UKo#fb;-3)5nj<eXyVAr7K30>(Qnd=1BxxL1N58N~L3V
z@eXbZ#Z;{-@fb-HABi`dg?@gOfmg3R0Juskp+bp#y+^{213i>mvWQY@!2wDdfRtYT
zc`HMZBl>hsD#N=|PqBzovdv2+9@*3rAdYZdo>fHKgjW;c-%!4Z8;cc|f?$ay&aL;Y
z7X}Fq<-7yuC507E(&PbaN`6F0c&6h={}{lTptI`$7SG<X-d{-o=sj;^c1Z1_Z)n)N
zBV`}kgh`RiEHRVqCk@ljYOzgol<YpijZ^b3RTP}tU+NcmeS9Acmc_UEZmym-bEF;G
zG*d5sT228yjFa=hvZLnI^YQI<iH)DND5%GeML%*I@K+J9ti4`nNe7f&j(+r&qo{J$
z5DCqYxnT#`v5dszHKPo$seg|jN-#F@R&+UneBgOpoQ!tJhI$L_#Zbk}DZ%dt>RQtG
zTHLt=D&J$&)ln|X$LDji<jOxD)mA>oC+J7+5<lq5eeY6rk85L7{Oi%yERw5^klxPA
zlW}->B9W~1)B&!9?!N$EsU#G0(6Xt%Tucs*X3x=A0fR`#nU*~4#vZJyQ7ZFV(%$B|
z?s+revE0PT`ulgUb31m<_MtKlu*=$`&TO^vm#5?j?Ny3Za-vst%sg@U3|oBFS@91_
zsmYf7lOL9Z^oE{gUQU-3^RMXq(!&lolRD*B4@G=z604A9?rD84`kD%4ICQSld~T>b
z@P~BAh|&mrXQb#1qJX-4l0>9*=qg=M_P9^Usa&He{iH9~ch1{=^~(-XhBa1vQ82;D
zp$4!&l~nhkbiIfLZK$r{tgPOA4x-2n0d!nVGyJg{b=(t)>xecSO64ric!zL@n7j1L
zy=v=r0LWq5RHD4hs9pfg<?EK^v_){c?K9oMP3S>{!!$xY1n*MZIb_x}7L6yX3NWOi
z!jn`Xua@P;u}_Y-*5VCGRnv-nz3!1fxuRfyA-TzsoamlSZL>uhKjj+gJxPNI{P8r<
z9Vgj4!f+EHLVoQUzsgtlml|~-0f3o{Ojn>{UBoTn%M>dMjvIVc!1MB#9mX?IS38Y*
zqc*VvfSJ5{u&5ggsnSeT*HhbpNpNDC0gQS0J^`Xz3iedGA$t@&Y=na$Be4eu4l@i;
zA5ADz={r2@6!q4E4aoVqdX5ZhAdo*O=qm~@NOPWoxZ`DUU)XXA3MH1)OL$tHDD-h4
z$A}L;ud#*S0abd`=lYTXy`wImhsm?=+MLm$Pb<b_<~0|vf+zlqbm&0>3<-jl*Gbbo
zunCrW`Dgs=RgwIzJG)6ysEBOyY=9MzM5Ra@oWLK(iO%<vnr~RWVtc&b8HrfN8J~{v
zjE&v7H<4H-=<6EtiKzr>ScMHs1Prnvvxb^0Rnlze>dugXkKYY?WL8hn<Fwbj0JM<6
z!q&n66$B5)Qel%o0i+%~d%tPRNmqOwcbA-t0mGL+&P{?gZPhy(0-X#?(sbw!Tw?Hd
z=gj&GL}QGb{MY6$Xw)L%aTnMhHgm+!TKEN{xG$iI2xjXXN2~Mw_}wPh&VcKZ#tv6(
z10baVyE#@2N+de#?-dH*elV?vZoH^KQLc4Os?k(3q8}nzXaDd!-lVSq2`4su<YpV2
z3l0T@w1<Q?qt87Yhq{_fBfeE3!=Nz1h=Omi$>{j35>UF6wFh_*D>=5WIILa%&ak2@
zVeGF-C0|3`&}pLTOvjB6nWDug{fasoH&A87<K+tZi&^<Gxl&$|N0~Z6DNMVr0~)v<
zHEOLPC-jm~K{lYT$&4@q>~l&}luKf3y@)n{=(shU=^{0a`e@Yeb)Ps<fsYd)xVh_*
z2u5gn{l$5);A`!J*<5$=r`Fox$d~RK8?k}N7=Lbo)Rc?{7Oe!uFxn}!n*)6Q8js%p
zr?Et`n9lK^<R-Xq3*bq<^2=SJ<mmKR?MtC?B<R!ouE6d^TsmeFSwDn?l`;zF0<T3O
z?pf!GWr>_M#u*Dpixd}B;?$5`%jst?gof#$`C%Ocr06?1&hH#mPvbD`r$uCcb-hWL
z^oQZWy{5Ff<+~vtjyXegqljchD$|Hh&Ek_Nv>wtKN75Czx9gl()UKpaE=wz;n9R4!
zpLs<g2_h20FN46hamZ&H4oUZ6=X{rx31|-|gd{;7y2Aj<%F?8_`}afcbV`7?*Gr;`
z_m>CgIVxRnaF;f+5~3PWDmxU5FbZSbKDM7n!i`3>#1%@3M~WVnAjiPepYH~V8~VfP
zP$OX3{a$fap+o}lkl&-fzDk&67Y#<(ZGJ6T4+mE`y^K%}?%{mXQmwqKAA}9eJ#H&s
zfy@D>b&M4rPV#pX6Nw$_$#YBUKrjn>J=TiUqU!NC*Oo6k^v8~7y#X%4M_DWy8{ga%
zq#g@sEkHCv97-hW9F)!hA?4c_Eoq}Wp!Ev?Bs+TlbG8PYpsDq$a+DO{MV^HpL^?#I
z**sdc6SXWU-#$>fq*6Q@SAX|~M7N)T#k#XJFDOGoJ&6F5_Qy;X9<vLXYGz>^g%H6@
zH`*EQbVB_V`?C6PN?$1)!;=D((gHTz!2m||=R2-i+v<z<NiD!WkW1w8zV?pu$4wmy
z{bobzN=5~Yx^+TT0bNFeCjL93(-~e*C7|0}haiic111gx6Ja1%XM)80dZiY>?>&M$
zMkOo2!v0^a3rjd`z;XHM2Z|WookI~uOaitayEexS@*jn?L47qRT1wf=&(~4ut9Bqs
zB{C@ATf}C#s<F<Y8j(RUneOT1<;)}*`n7mpPPg}6+uwF6T4EHb-eUzY-~wDj^7q|$
zhpHHHd3}Ckw~t|_QbO?hbgjKQ9}Q?(+_;F6scwT#^DPKRG6>=XH&&a3SY!CGV~FU_
z(yPPkkjN*-!z#wqKNhXF?399oVe5diWU@zwjIw9Y@A``YL5kFUI@&e&^a?3hn_Jof
zjCO0l^lo*L^>qaDY<?Iqpa6pamsKHF7iWv)Bc^<wGh<Z!c!*(83Tu*92H~X_lUnyM
zAWp@w4v$~wA}X}~mJzpJ8&aBfHvsPxT!@f{3`~s`Sxco=8Xjuox#(%+O1X+c3rN(9
zbrce#K$?Ltrx<tx!G{@5w&C<{HSlj?Qe294pNIj;jv=VryeePLYOz#CJ&Ksh{gza^
z=Q13w{;XG)HxmxL1t-z>jk3pJmwThXMBn6}pwZJrn=`IjMNP1fi0DHvs;ipfAw|b<
zr<O%nUYY_kpgVMwCHjE=bbt0-tlVgN#`%3D=7T8z2bByBfdI}Jcmudpc<CR7`0IXy
ziDb`<KP6tLS&(4y;4Z{~-5dC%lt{o4L-twGtU*D_UI9QoR(VU(lWAJ@nZNuLr>7Qa
zSN{6_$<e`Qa+M|Q|M(RWUQ(z`nX?m+oS7eznleMHRMzt56BGTgJo&Ku-V)0Yt9H=l
zVCZracA1Qzn>rozD9)$6h+L+|qrLk|n`n0(vw-2xxOB2%b?^xcV$?cy$8`swm6J(}
z;F{7u`9{bp!3W*D(a5r)74y~-F17F8Dwn1TW@KYvH~x~Rk?$gcpR**Tf|l*^OWW!@
zzXhQdr=+iyRObSQIKLVgy&_XK>(j}62}b5?Mg*@d-5~`V7X_svdbMgUU+U~>T8#dE
zCL|3U>m4z0Y)pO2YL<yByAmz~t<u(T$hTH3+HVT6)6kn%GjGi&fGq3e&A%D6*vOW-
zHHn&!F}j^k4cLHORbuSY4mH%j$}x6vg?hsw^dDRp0a*N-U=<)=!ap#^&l<*TTwhLz
zz$FiWWShjs;@N<^MWPJ+S8onKLD?47q*#+L=1u<VBh{k%A%+qDaWRo4&)Mkv8?^uj
z+b%>}@VWGTs%1RFtW}NanR!rr#|t}E04P@&0}kPT@0w5t`D#o6XIul>Rl*wv3o4s?
zF}_?|2sN+0iIIRIjp6_1O*#|O+_1=C&l%9Y6;@h~hhV&U5sv+gPWuq6z!~-c$o!pf
z?UinDPT@U1ipVc|`18Eg5uRAks&T@a{CpMOtJ<X>=H~SEm2j^leNsm_@1SD|&U5Q<
z6U{I1w0fo7poKFMa0=g<nU2|mM#CL|>D4Qgs$3*SDa{|+Qn(IO*uqnVWtTUP02Rug
za0{$ru5z*L&3khtYPP=BglY9(ftd+L>)cDfZ0gExVu79Rt-lG<$gX~}zMprwxRCnZ
zOF1Mb0ZnhT7`8tqa06rvfGfZ`F~sSI^yMvm1%1S?Nr96EM`O^;(DS<BHV_wa7WB~B
zVgrQW_!?1Y!}YPU%p9r=@Qe&$J^vC};YmdM+Av=o0)TwlKs>{`FY{7ToVuG*6|iNT
zlF|Ic&qAR?#Wq89Irpgwaaki;`wh&F&&0}i0$e}iQCHJ5%jz3K^N?EOPf1hPu)ZpJ
zDI7yg*9985DaL#u-pGI`WquA1LdOv=;G14nxx>TC8VdU%{Vy+4$%hX<b709F#ewsu
zUzM-g(x6I-T9hKKU%-|r;!Q=rofwFXC_TQ^0A8AfvrdKSj!3jdL>BYx!ilLtAQ4S}
zCB-N%6qCoPf#c4LCwUNoHl!^g9ug>kV<Lnql=rw?5|odGGFLk>e*hFl(3BacLzBTf
zD}SPf4l?J!;o>yn^0FL6s}mw7I2!ULfa_jspf#Opc*W0!z#Q0x3kylAwqpT|Na*}I
zGu_9y7wV%vZ7^Ucg1>-X0vkPyX|pj^z2<XlfD+mD7vvNe8pdb&5pNUrtSqQ#hHCH*
zGOOZBVsf%sN~Zvk>Muo~jlJ>(UJ7@#KezoEXgD>e^V9zn!vKdTog&0hf9JCl8qYz~
zwi6mJr%0_-nT`PXE<6E{+&Ie%oy|f-8~8?mu948O_+DsccW_+d=MhIGr}*(WetXgb
z0SE0NJ5N=(kf(6C*pG%G6jo8j2%XHAUA{kcrj&B%=ty4}efyp}(j4sK^WVMuGc1$D
z!gzjaECkDP`P~fAro#z(ahtEysm!*%`UkFLO1>a_u2b7|rP~~#Hm!1ERc;q<xAf8Q
z1Y|+Bl~FRa70YB5!}rtt*&e}-u}VVenVRfN<HLaP+(0V>suO}UQOtQ@nk8tV=)MT3
z;?K5$fC--03dj=)35n#r_w#5rN6~SBlz5NGh?CUy@sc4*+4hAo*Ve=+D+;dJ^=kGY
zkrBbs4AOUMFZEe^$v>1kiNKBDVPGqBpQP!1i9-zt2sW?Q56}O~>z#4Ch!?Gxi>iJ%
zqZkt-D5A@F8`oLhjZ?khW_L-WCu>Ww2lw#UZ9mQ09=ajcT>?IMD2ad6>730x`32PO
zD2^JhgHx8hdo|LYi#z2XK})4S>?MJd+luT_PFRyg*0%6{q;7j^{BYAJRAKR`MKXlP
z&79u&?zvaCME^0b0F-ep4}lakI1Y?s)R47hR}+0sjpz+e1t2LZ=#RGa<W-_>0SdMb
z@VsnV(r86nxJxm4^hyqU%ROrRi-dmbVNV*XYpT#K<?lPv`n`B5hb88$AJ_L_7)QA)
z5M=>$zi>hX{gAlY&t?%JYb%moYKf=vQFGgy*b4(q;-{l=n2WYu6@9y7Q$#~Sy0X^@
z{;^`!MXN7_AW2fbaPGX?^&)dZOp2CFXP?Hkh%=?!qRwbId?FN(lpbm7_yE{fUmrfe
z*rd7jbm!^x*}e7F{G{`FJJa)1)#Om@asc9TwMku?-n}^V4$iIq`Q8840RRtC`2MSY
z$X+QDKhm~20cTe+APo~?`jK4TP8K^RtQ=4<8d?v~vi&>!hVklM7UO_k$HE)@b=M_@
z{5oQ+T;$8gf9u7N>BdRxH~awG7x*ogN+%JtO_lnxJbWUZyV7wd9p#kiwf~Y50AHQb
z4*cJpfL9ysFP!I5li>Rd&|98XHTfNXE+)U6b4(@Xeq)J8zFO4SFNaM>jQ9JhfAIOw
zXC2Sl&GZPp|ItZ;HDu1SG9+{XCcyjiKxm{68DeF*yvIug_@UKQ7UF%AHa(tN8M70E
zLl@2^K!W|_SN@HzA4|$@Ykc*{yvRafwQnIXzBnJz2oSX`g!{cEVPP9#2Si=*NZ*y>
zsU3t3^F;0J4uHj~)9P3LXW{kEr&HG!18e>s;Ohrckz6#d`99Gvjn(z`0Cs?zAMXib
zul5nBe{Ex>Nd1uXr$}gi<R)~$cSit7AG87)%x`IxR^x=K|7tsdfVBxnb08zHrxid)
zzWx0=$b$hX*jFli+@wVM_?P{Kfp_>+<MTzABUO4VU+kGu2;Z=cc$joAR{a%k0l`@A
zhHD`|wZ=z2>sANeAF{H8!0DSXVQ#NWMHZH50iPCn()EuKwdcuqgW?2vBbU+SjH6G@
z?<~@uXJ}O906csGzqEli#_yljbc7S8_Q!*szE#V(a!GWdG=zNE4P6N=Gn7jyzkark
z)w8}L1)Bxio9ym8O%%HXu=#bLY|##c`%3|%1F&E|aX??M4jiLs14#$D@5n~g{bT8*
z0j=8YAM8I$Mv0ReZ`^Cw1J(lpblji+?4$fDFRz{-loS#&$fN|ZSAe1y@1yf95PW_s
z0Mc%yT3BWM(Z(tJNjjK*=3=0ZfB$qmaY-3r_P-j4f9@^tRRWBCIu!^O(r*?eL5q|)
zaHU56#m<r_KIrdPM2FD+%dPcnDT|}kh412Kv*@*7+ubnj@B{Mo>cp>)#Mi%Pk|eXI
z@|w3y`MJc@|B5ybl+<Nv8IqAWYXmv}boPQ^7`nA6czYWGMq&`jjsn~ldhqFhp)g5Z
z>iuvpWk|b_Y%L}?!0E2ciwG;E*o;;IVRhh8*$K+9Q3g7A1J?}t!A?HO?5uaCs(!Y{
zr(s;+Eu1vR-E7deJ+udcUR^mQHtu`7N%d?51OU`dNHGYkb-<JlibqGD4tQ_!pztLM
zTNdZ#r%xkQFT)aP@HdRuS5{q4A>FU7Fg{TV>7-pxg8zDqa$5-2o&E*b{!!2YpWomg
zUKIFkOD_0#e-`n`iO<>e7>_~QpZS?D!a9%VgUfvp>5V)oXNKJfCs#|P>$}kcydol%
zP03}Yx!)oSp_7!9A;2hNJ0=Wl2Y}ar@Yp8T(rD(}$>NVfIjdbY@a-v4+$0kyutT@d
z_d{UvyPYMo13<gg^%{zt+{eo8Y$WsjrsA(85uO7MuL=U>k&+KLP=B!lgR<k2Wpz(R
zyZUn84*<8&{($&;OpU8l5o;)B{Mz(yR8M(50}~$doo&tq5Ofiy9we+^>}~c5cSwRK
zwsuLu<BMeJ-LHpNfO__y=%Ig9X9JMpM<5bZop{bw94c#U6hr3~1|8`({&Nc!zg75=
zoqx2UZ@giUHsBUn{*CT!Ch)e3x0a1m#LxX=>Y~XMKk!Qo^5_s;_ZvCdCbEZA^8zi#
zJ}oqz8%w=#+XB1Yak`yT?C*Y@5OTdKD!fd4j4M$T&Atyq0a7&2dioaGoPY#`!|j}r
ziV+llP*qramRA@zaR3}gSDJrtIRJ(r_9FE*#=LisAv?6momma?D+Ez>oY1t_8<6-1
zQ%rt;41QmXOpHKnt__3*iPo=XFiYD3F&Eqg;~b{N+>+2GLn<%Oona<|Yb7T4uU!Vf
z5cu4azxdgIa|8Crs6-Sfk^e+s{3E6SMz(MN2%_|PKsq%%+i3*{fRfhkYKL<Hm@$Dz
z%f;!kEU`}yh?u##FdZJejL?;6^qD^d0;MpDYtIeuu1Dse1!X2l5ZNoh$0*E}63}x5
z<cPrKlu?)P15ea)RaJeNAh(zM+TD*740r}cN*tplfakkRYflY_wF?cyp!cy(g$lm8
z9%&9~{01Bd3wJ%u$1)F%^#mjzq=VV7$q%!kR>lEry>W@UXe#ibW}Kg3FU8ePUL{m9
z3j>gx<#^rNQlMg$`iFHq$X@IA!5s$-AwW2}%%KT<kWe_@G4>z*!vAV2eX~-ozaVs6
zMOyJX`IXf&Z%;Tj`zDkVz)|RtMfNQi#g_oAaAOqWY2#f;Q3A;nrMI6@^KRG3$kZk7
zlP<Xc;0wUo*m5wR2E_XSZcS<2Om!6?PHw+EJKvnVdt1YgA0!PbEvMq3csMl3zf225
zwy6He-?>mm5B~iC8onOk@<(l<=O%E;px4F8+;LC@OvPkG^mU4yWF@2R&MLK7;9Q?|
zASLk7_82J)clD=&1I9_Lo#E>a#Nv;eV9xJ@HwPN@f8YZ0faCR(1X3~!Ui`rK*keia
z)ywf{_;8o>`E8z$qT++`oB!`|*UxKZcM$)+#>BblH=ACF`lbY=8J&*3duS44W+RBP
z;wX;*#0j8ofDpSGFr99~UjVYr9-J#!sEY`ha2Y15`&+S{jk3dC>}?>uDsQw#Pf7Pp
zygvj6M+4h9c~3(WZ8N?;Q9!nR$9jt^WyY#fde!NJ`8S|g9A)PN>8!xI)JKi5?p-hO
z_?x#{C`JXa$hfCI)+!ar(Sx$DwDMa@?0Y$SxL6hmP7-|`d#ydYC5tak2nj{HSA!K}
zOAHgfp!5ZP)7}Jw7TdfW5Y7slg6k-KT2x%svzizwQ~*_b4i=E%f#x!PSO5sXoZ}Ia
z-M}?uB;!SC?99czkK#y%wKU@d!~frBV~zpLNt-;dAxku$(p{+oU{fMO??~*2+~AUr
z0^@*C%EDure37e(rLVt_c-eA2wsh)6{)}0%OWe{S^oV*tkK0NbCl>f}0b0?)N?<w`
z)dr|vz~@hKBNbBpGWHsnYOR`?9n4aLVg6Z>k#`xZM{~F!v29>g{+LF6%Bg)Z%FCvN
zej^=glEm9Of<^>ENPF`oybdt!W@gd(der>paYKJI5DGgTO<bA>_x$lQ(Lp>!P@U><
zfD3*6MB*8A!Ca8%Bby{ih=H8t=e4&csF%{pcLzXqsk#b-2syYc0{#b^w{qV0)n!<q
zp6YjD1??hXnIRw^(p?PeF;)gwO!<57*n``MH~`#l9Zz6Im96%GRTSFt@Jv(uchf))
zz)o>>v9SW_kbMoM?WpjAdz;VTW+YQH)so#&TcE(wvT+(XR0~8{6$>^>@<%lRZ~-8*
z*7pM~CdS+X!GIHGoOh`aZ6tNlsU{)@2=y;46k-0K*OHARg3$I8w`?y)gn0q^uJ$v!
z=c}DSUC1M)%Q!t9(Sez_LlXp^4Snx;n=C8hc@N+ju$@JV1$TkyBEVwxTd$%(3Aj-w
zq^WQdoR{^*vS62|hwWE>w$`W!qD}8g6U1Um;6iERDo`+XLpZ3N-|FJP2vTP$wnTI3
zXa36aOW^a5!L9ibNQG#=H!5n0lFO4VFz(^_q3R=6R`*bv@7TlGLN_0~5oPckj1Ayr
z(sY7L)mDchHggkzdeMltO2XUCQdRbUxO(fjs<Q8YST1#GP(qOIMv%@+hjd7bB8_xN
zNL-|mE@_Yk0j0YH3F(w>>5}eeGxM4G&hPo3m*6>Puf5iK*DqUD!U3{*`v`1vaddj3
z>HL*hOF?Y7Q<;N3Aig_DTd)O=|DVF+o)J9)E@5LWV4&E@aBfrF`h-IfpJHtFa7KC5
zR{r$LiQq8sz4sMYyO7xkuahq&h8>w`p~(L(Gnq`FvH`Uh9AoZ>f_L9!t}dwKKlGz3
zIJyPOJ;VUz1u6HJ_e+OY3?%NJQh8K<T}Qoyli0I4;Daj%96ZpeWK#N%oyhO3T=WhS
zF)`(2Eq-MS1UT3rn)2tPKm@q$6Wqx26tiYR9yEAzCOx&P1`>*!6UhR#zZ2<-wTE-g
zQwizc&la63jcythwjYh^<n#Y`mMF=^+nAgBW0zQ{$>ef6u#k2=1Qqw^N?dSFCfsj>
z`$H|xcYhXqt;NKXaTT|`^7iDqmz~t7L*wEMuWuat?|wXoAcZSE_r1dxo*@0myZH-4
zd5<jStD6dX=R@HzZ-c_D#^Yt$9;ZjNmGNeaGT*H=a`XD8OKs79T<3R%W&q0a+pLG_
z+hWO|HQhmFW1UMePEmOaW@I0fj-XyA3p!sIXr;0gocHgc1G!gycaGn-K*RYnR$luX
zK<GrWM|Rh7EV?coza>+0wu}SSz~Al$_l;)Fb_5IoEotv{A#kDzMBIFa_^5`o*M5WZ
zA$3oB4U6S`>FQ1Bbc9u5U8m?bVkf^?K%B-)!FX$J<|z~7{SFX7eg#l3uxl&`FlOz0
zH1(Ho)mhj`kzkt`4K>ru@>+X+#o*91ITDR?s5|Tc{3DXtlw^%YD<5X7=O5DXx9PcN
zRBpzYL0bpzHW54^RXob3kp{m;K9N?m1cB!pO{$T>ZxL6h|Kl{*G(VSgs(oQY+m!l6
zK<U4d&7k`3w+QTl9RODfMW6QoACM!2$s4ODs3?a(tz?vY<ZV>q^(GlJQek_7rw17i
zI!SD<&`{jQT_{~!9X?xjN-LcEL`>_yuY0$iW=NG|W!rY7^x0}lt+jVGHUn8@4=f06
zX-(v9*ta8R-NEtch1n{wj#Qqv1<J*U$#tI^12DsP0=WT@)q{FoCqflAGHi|!uM$Z|
zW6@$%O1x|eK<q!K@WTZ*Z4MkI<fI4T^@DUbe?Wu0!gD>X$ppd2m*PJ9!M(YYNvqnM
z*&jHWuK-HeomI&oko^ENKE)&~XL<xvz+x>Co35>`#2g9`v@YRCs7a%LPP9XhN6<f{
z_xE$0!?kd6n|_n+FWHDm?<<w9X7)&^eW~rd)xPaxnmjfZKDaeiJ`#e3L3T@)u!u5d
z&}mEl5Ebl>IO?k?bJJJ2J1({gu!9mcYmgl^$Ix<<oBo8VGO#eqGl~5q3h;b1X#Ssm
zyi^Esayqsk!v8uhQr?1fD4&Sx+|ApMRz_B4zi%cU(@m#NjGz;1GXbm!Kz(!Glzx>1
zh(&-EPD!U^H4IwrLBc{4`Bhv18Sax&Kft4k<1^RwRlFmk2ORs{^?0D%>ZzYguh%gY
zjiOgbyBeYG$*`^pJpTO3;S{t$-5)oW>~QA5KJ&+UvJ|~K#lavrOup>u`}GiwHk%J(
zL+o=Ykf4`A0JwYx?B#B&^;nb|^IW9$r2b;tmg3xV;eg}UpB@NZx|#O1d8ZG$P2abg
zOR^<aI#b|>;KOaSIYd$+%uO(uw05O#$yW_n<cF-~aJZ+=H1LT9c18fk@TyjwTKr-Q
zSjY=J{8UoO8gFghp+}pD@k^W^F{itO;|t?BFB1!C*9!j6(*AEv2?_MxqfV4qjB0as
zk+$(FThn9`DD=VamhVG8UjOyQ@1fM=;p9f3uJM}Owt3WB0^H+W8{L8^iwklvp=0j#
zidO8I4Zn94JnU<IBakH&AGn4NuTjaV(Gy*CnE9y#k9ZU*0m`B5Z2#Bj-a5PY9gQqB
zS;$EejC&wk0;m-hQt3U-6Rv#)69bt_SpK4b1H0gu@-P5Y%?dESHhVQvectz@p@!!<
zun@2ti~-#G3SMe9N-Aq5scEC_Zpq(ZtaN)j5~QEsGtx472*s!F|AB}VsboXq&JB-{
zlZ<lN2)gkQ2NRc_Qeu;s_O~_FAHXrzZD4km&larhc=@o7xRwP9OKMUF#zX$|kbt9@
z&0;MrSSjcs!a*NPKc+a%OEJVHfomyE8ArBj?f7ZT1xz~**ZCN6KqCiio!b|eZnGdr
z?3?a%uuL$9jqwCk$YwTv9yviBiLN`sB&PldnjA3X9e~q^`x61KN)a6r#7H@X_&p^Y
zHq;(vLJHVapRrI4OrzK(Df4)XwtZp1Rs8HXqo7jsa?j#W6V^uuHXEi^2H_!@%TLWl
z;`9>d;E+f1=m{ep4pmeUFTqb!gjzmYW}I9CjB?Y&ncubr>+-ce<^%E_lIBi+`&jsZ
zJC<~zllY5+i*>!-7!fA7a(KC_lreh;K}t2vFt6Jmu*C|t0;cEH%^DmkbAN{L7Yoca
z^#*yS6Y9s<sQ+8Jz(oaBJ5ga752qQ@AMK&`y=E6e^6(_oosYmKsz&3W%0%YRzeBn4
z4etO}pi~~PPkxLDP^l-b7G|DEeuM9zk%dc6E!46}2It8s#{nS^hawOmS)$9<GMOQ_
zzdeJX9<1ujF;-nHHKl(yM5su;RqTTjR}}(}yH`o>0>s68raA%8=7KKZ&-opK^7Ra@
zdQ((8+RAH!c!ceouks{Csh6GCRBz+=NE1_Rb&AdR@Qt_!_Q5Uo!Iz^&x29yYA>XUU
zVuPq~@eBLBy#0RBaHa90vEwLph9Tk#O$<7pEg~ff0vz-&pX9$4Kj3zzz?qXh>fuPf
zuGVTre1$^xYLp*xvsLN<?PY?5pYp)fafw@jX(_{bN;^JaFOSIJLWjSHJ)#r;ymM^8
z<%?Pg%C0%p5cswe>E-X=Ey%K1pVe3UyEqK$WWI>=G58G{4F`|rfGlzOz5uZ=Lp$25
z-197UAfJJ+L}g6I;9g)s2|Uh#{L-jSEPj9{RHM5>xsP{b#_f3v|5YKY-oIuOe<?7Y
z(+q_UX>#R?TNITXp*%&r@Ohc5mLFx&fQLi^FzNCJx~b!ML=!OHN+;_aOi8RUV8l6Y
zU!xAi?S|OZ_F!8c_4al<LlY$_-=joz=rQQ|@{-E=yszvghWhiPjGYa?4*<F{1EnvM
zQR9E#3*&ouj5S~J-70Qj9`Iazl}Kd`+7lbIS23PCB-C~ayl(CMCYdO4(<F&g-Jvqt
zR7A3r1~p(90k&)46*%BL@1)^tDZuB52FgsZd$U=2di4dEvy*8dmqaRL=LWvz1%RCT
z*W-^0S8qxPaju6&^A<`Pur1LPWtg3S<fgt*h4dvSnD1(p+?-0jy81i8DBU*9UBHVS
z^TTqu2hjd+-2qxkf1>ZLN+#niVv}2-gOB{f>5@_G<`ZGuLY>&`NFq+zDi8J~U?Y?L
z{-axQlaN0AU5sE=w;20TxMewcm=>r4O~NkFZD5-)N+pR6t<+Dw+2KC~7svCnJn)w`
z2`HUjP^()3deN+?u*5G*bmCJd7w9Tp5=ytj`5^~S{$q|7AP@DO1<rX6AaB8xAE~85
z>u*ug91X5X-`5;{?FTXRgkb*GF;1W>&aA$gT`O<<!^SrGz$Bi9W(WiY0C^OEyOpuz
zI)or9e*jQcfuV41{0zWZ5Adce4x*#my`LOA!|4}QYjpqg%se9Ur(Iz?>UbP|mWA0E
zV^PN22x%3t57w5MB>t1yt`4HqJn+U4UI2eK$$=ml=#|~jhkzZtZX4O<qerv=zAP9g
z3L6K#G^2)I77h3Ro+zlE--wq!Rb79y$Nd}&_VI32n6o()<x@}Xv>gG<?^I(pbSQ4P
zS*G3}yONEnV$g%kW{HWNZ8VptA6uEEQa)09GIi0UaV|`u*OyRjtC1KU0y+}FfN~@2
zeL8s@pHbN_%)rl7XNH2jYVahQx8o;hZ$qi(9@6%bub7j3Rl}JMQC%y=R{3u3qVPIZ
zl%n)Yc8EgaF3rhs{?W-{<lxRqp$4zBD^@867}Md!TCXjG&Au{m5-qwwJ3j9NgZ#e=
zaD*?2I!ctLe6=+CM%qniQ3)>V@QL~iY;eH>{FU_nr;jtJKUVN;Br^ff2}nxcpHrzP
zU4gczJqjNmwSIBeX>t*%(G)s(j+;)r_J+$uXVJcAK28XgZgJdK9Vkm{u!4;$8IU(K
zNuty_aL6o117Uc*2C6Ua!aXEC)MUsVu6Fnx>{ON$K(Ye3p8m3<26N0aPbUaO+|4Iv
z6(3P^gQcrE)<eb{&2`G7W*t9)-c;9OkdL|Xd7rLHpGw{wwQwh$=_|~f?^GM@r_Y|~
z7BkfRninIb#v!{y1P%Z&-UEWU07)9Mu{ayUI@BOZa{UD~ee!xsj_q%xlgOVh;OgAO
z-e1x$`u~0?_wX7EcqV0%{!S8W*U9K^sy_9qClk(K%is)4<741Q8@vF0QnLMs!`{Np
zijTwOD<WyiR!o--FUIe@w_Uf-T*qZI!Lq;RG)!Yc(HZ;Xa7={n<;O%wi&xV>L-RNz
z#FT<JaJwqOF;-HL#|gfasiyB*zSNXh@hX5`HP5y&j_dipgIv6!nX{bTtlw%ftnSiY
zWDxb^G#gq``<wowiSoSev9hoiYvi^s$y~d~_zSc1pVYn(%DHBD545$ZWZs7!lB&ka
z9hoGMsATYs4l41iQ>q4y0A>w#(^27+;A9ox&ix~IW`!snGaMWHoCUcN*SGs*Ly>HW
zm)^XeQ@T92Py>h{K0-ZMQXv+rB9))eSP(*naX`tw-YU&K9ic^2-|+uE2@FsER>6i-
zNAYV-zA4<uW%AQfQ$V#4hb_1yd_9w~Eb4rf;VpdyG`N!<@+-NG$Xk>77|XHF*qh-q
z``ffWWr;ZiLLROdF+R;wIsL;zIYufoq-7p1$*{QT#%GuARXLL=mM<(37b{OQt-L-K
zt@>MPq;6^P!D++YUOK6Ubc{S&Rqp?l`?$t1eN=G+NI7=ZbMdFoe%27bOn4vh0+^lW
z@&51qX7<dWL2}n8Jf75s$bzKRQ?f+y%o8<BoJf4-X|ss&HI`Tmzy@9<##K{Bh=dXI
zj9arq;38(obb#uT$IHj>F7qm;Wb(}A(LXgI8;YOt#q&-5Viw-vLRrk?BkN)6tFL$m
zPV1Sq`!7Oc(Nlt83tw=3S`<<Z#&noE4pkzQ_$@{RbNFnH*dE<|RY+AIJle1+x2*{1
z?VFII#9di=^KdyC%_W14@Me`=nZt5Xk4(T+&|})UFy`yfo9ssqV?xaJsYQLhX<q*%
zk(ZG)eew)k<>21@N3%}|ul`{=FZ$Hmi94ysH_thvIFoG25&e0au+s-?w2na7%C(}~
z!>y5v&FPNcenO~K4<JgjX3bu(YXqg29xu^&UxU8yrYd1IYrZt)pD8!l7dE#ZCa6R(
z9kkzu-hck!B6j8B4jyAeU#-$-UBc31yaKYzfg}vfCvUWhKWk2t-sFc#pF*YHvCWrZ
zj$1^IA1^w>DE&{e^!j_{Mmk=1k7kq=?9Q6@^1aPWcRHcqcnCZ>g-l7x+<36@QwoGk
zcgYj9%%nnX9?3Km3sj~KNx4*$lP3hSd{`xk3cO)AATD!L54z6Iq13@U>42YFSFM?!
zr{cry=eC5h2s3h=#4@D{pO&1S6E}<E!>>EHs)jpSL<jAc9=80Pz#AZkza;~pRnj3^
zEnuVdMjLN4$+AhPc%TqclxZN$|3BD>qb?F;9cVCb2i^t<7k;h$*cX8{kTpv6YCk($
z3JffxS^W1IYfo@h@GO8x8w@fd9S}jOSmhH}X{2eUKFAWc9}Eo`2?t_$x^FNYNijMZ
zePX=uOp8zoYISD2x1{DlfXmp%H}NVx<&m>ROu_RmA4k<(M1KO2PN_6vf8^6-Al#vU
zx=B48xo7N)U>a|Tj%5>l?%8yljr_R*wQB|!iuC7DqP%#)Z1d_!D%O?RBF8oo_A2M)
zlN<^Td@7t-?<#{v{oJe}zr|+R@O2!b;5`yj9}rDp{I>mhEOJX}{Ph}>mvx@~4_o(P
z&Npgxn==H<S)yD|j#z)?Uwg+(W$1SyS}dJdLNxbls$xV_a`yZ1PNo=@IOwHeqP1Lh
zibT|2Uu!TKkW(d<>?y}io9aGiHFNnAI;taiypCCv1&I5C(#Xe$<ru_)S3UM8C~oq1
zQ<R<G7DFRGKp-Og)ZpI+wVa(t0f8ejgKqtH)Qa<Ag+ll#)PGy@;m?oO*vu4UWRV&&
zTG~>~4FeA$_?F_2*(WLYrz=X|{0n5<Bfxd65b@}uSs-(sPh-35Gz7Z-9>niRc8WA5
zuwOj3#haWG;zeMPX)*W$w8+!uBI1-^(M4PK=z;R_H*h#CON=t)P>eAliUV)$0j;`K
z(J=+f>X<C$bNM7Nrh<l-Vn74bKrfJ*5%hwGn~8U01F$D?<$2Sb+du>1=HrBMnh1o~
z|G3^mqqlY5mrqQ`P=Xo=8F{+88U0D;6)YG>6yZPzPcFmj*TIK|^9HTuPf|8*YT3e)
zs9fTyX;z#MUbTuzdgX*F8qVW=k9mqL*BT1*5Dz`Qac^&%H_^UT`_Q6ua!R^1V&AzG
z`{RVOmjsCmv8lfA;huU?G($t*Q~NM{{G7qjlq0MR4m_-*c;<@W_0HF!T|a1b@Pai*
z=6zvjm?0Fq9HYNCqMcp@UzZTKtCsveMjmDKnVJD?K-Pl~7u`%W{Gk5In=#=QRM&P<
ztvz}Jt<Gw{kj2!OW;S><_UJ<7b9`iE2r5Yrmou^~mM0HM>kt12b?n-a)0ZKdb_2M5
zbH$7zuMN!mFZ$6PB)A;>3}I<&cpV|0aSxvoe;&z%7V$X*U2h(u8FN*4VItMrS&;VT
zB|a|i#RNCQu?{mAQPiVr=WkGa|Klpu_{i{xAbncz@?cuCxX0KTT+Mtc=O>QQv`X5p
zBj*<n5BIk%^7)+%Vou`=A@;+ZBNAGk$&FRq5HKCSL$4qOp@XOZ)>FWnM<AZ)$u2W=
z!y_O0y>jxy&Zs1gN!m@!@DAb2<bV86h~W>&0Z)1SwIzmRw<8AFxsOKJ|B468H`(Nh
zr?j2hswnfS<k6b|lkUdRpeHJw&IqN3e`W;~FrW7meo-c{Zd>^QuqBYdP@yUBXa}tR
zLzHqsW?nUPX%OuO4+b(G;IK!K7)4#7+JXMGl@=cXN|j=+Sx;XT%*!nRjn@UEi^+0I
zR^#Vqyi&O{mwyZ!{;kP@u-3qVXFxa-UCX4*V_eMo4Tf`jVVetuRw{kDT<GwBP<SbU
z98$!gmAt;I<($kc<<N1NW)3~I<{^mhSo$_9<A1VH_7DMxs+oNe@I1UHH;?ESWmRxM
z*M$pR@;<xx?43fV%cD~+ndjZ9gqx#k7pw(&+U`=7j_^S2%R}I9$FKVKGa<ouGre!O
zO?B;@GbX--Mp;_{@_wLU(O6e{E!h8D1L?_%X8yqVYs64k8~RDWb6PTiyW==tG>tG|
z$dw??7?I9vJQ)XbjJKaiERqvnm8vFxevmn;fizw6#l`R0O^>19bZMj-+jQB3J4!Nc
zev7c<=GTYzeo6)`>cU1wT1uq0u?4D<W5kn1DDYz}D%?8AX$He?)C#)l$oFsapV;y9
ztXN9oohuV$GAUjZ2q`6a1KI&tkoN<T-oXg)!9DLcWZ|so6Sf39Sm5JTQL^qdelG@f
z$jxgD@SU#zzm|y$a!+A3`yK1@Ml8Lok_s286y}j0FV)I$j7Y%yElfEf%5EuaMpSip
zxf+dt@~xRZQ9&o_7F$b!GG4GpwOGOj(phS;LwP<uiGNxrCCSVi!G0m90H?FwpUNax
zwlVCHs^p@U;2*bG{f<mMZ7<&cOy=Ua)*n%(e~i+7Qdh##W4Cn4d!%LJ(Eu?8rKhcN
z`%|OM<)_G1XX|JUeGi!{oOzKXOztk75#(O|U>E3rHmiNQQE)y*^+z;?_JDP;;6!*#
z`|`{KLQsuw1<PERjGkbwN5kFvrg?ls!l|=yPCTmpv-3k!u0Rl$(`mQZCzT=v_wQz~
zOdP^7fgz|EaZypvuo@DiQ!VxJ8pn2>@<)w|HPSr>PtOR|*b+oYTOC)$M9)*BsRpNj
zX;MSIZ?Avb3b4~F=N~B|=&YT>9m39HQK(bT{U0RdrgpV_podr%FziWG;4)6<c#SK4
zW!H1L{<1mA^H4ub74aSy*?W5dCU?&q0bleKs(~M)Vl-t^v3g_bH#rUcabRzZ6?A<%
zj?Reuw!u}5{cqm|Ot_wQX(WtSfKs*V9}a-#C@y1h-vPrW1_Mg!`w9t6e(@o@NSH5r
z64^TCTA>4)noNJ~11<u1Uj)Suo5Yi87%mjV1~MU4ew<~_o$#qJ*>NR~Y7K-r$m>zI
z@v%tJT%QZW=<yfW)y}3F;pPEknjfd=C5=Bo$DV8^OdoS3>M|&Q&8tQ35hocoV|%!L
z=Jx>sE^0*mS&4DXSCF4uoe|PW|L{6<lOMOcH^FK3MT~}xG$nr<>P#u#gkfg3yC=c^
z00A76ud}dKu0moJC-P$kAuEf3IBu|arnGTia$-I-Z=af}Un27vk>Wxb{@D!cR7_^k
zCUV!6GdBa}P|Z&{9}Djp<?0A~Y>{Pkr!b%mJ8hb_2t>K618#O_y^iG=3LyJ#>`85L
zv@OcvB>-DSF0hggFG(ALkD2MxSryH{ED<Y7Xyp>@a_ej5wir&;^M7WyB1@>IbKh5s
zmd`&Gm(Geiym;k5UeJE6mdzxhr<2kT+Edl{f?c$rL2?!)^9v!<A0nm)i#x^Y!k8d%
zgmxDa{`}uGswE#aSIHctGC5~*E``bZX_Q9?=4en|Sbs5~;tUOCOoPq`=?05>2o=NX
zkfx#UU|x6$m%?%nPeMrCaijvfVB!@A|6|;Zx)8KO*ZhM)lpgQ_|Au&ZA+^Ju>#9<p
zy$U<NahJ$>0Kes<5|2{fM%p<$ada@wBA2_E^a$aVeYe$!_Nq=;;-b4i59#u|<lAke
z$2qFsL&{W<b1VrVZDs+C^@Bt=74SY3hpCtLU$=u{0Pe9<jQwG81ULi0t?DtDh4tIs
zOf};l^nnaYfq>jsF^9LmfzF0P*JtL+Gi@)5H&VG9krBwjf6q;YvOj}f&x&adQ9`Kh
z#K`MU)Hbo2otJz7Bn6Cx_xsg=XOuosG>V)~Jx4$|vazuv*?oUgMTNvH=F)t9+=<?C
z#a8^xocaBGj`!zF0elO>IG$&7p<jcan-3m(>MP4ekS-0blKeY-njoR8;c3?T=v;r9
zLw_z{DfyKOcjNQ++a*;a%e~hRzfg$STz{AJ!!<~ebZrum-5_M;5LFKKw~`k^A;a)j
zO=qvSO2e4J{QTe{=2<L6?Jd;#vJsq7<ERH|Aqds15{IX6sdNCix-Iye--~G9JtNTh
zVZ*FWCaP=);*T#HnkhA*ngxhtcOUv6t;-ql!?1V1t`A{g%XC4)4`QbRx1E#dK9PTE
zbs)oqiW+@l)p=4PXe)wSvP5a;sJo$)8Li**fVRH)NU`nwf#YijSHzkFVb|5TLlu=C
z0O0}C<eqPb(I{AC;@p<`qz`C*QpUe!hje~IQxxUlMY!dd!N7JU!?U>ETLlXrK)rL9
z&IOAeK)-z}5C>N}7z8jA<c+uG)Y7I<te{c?`%EG;7XesLzAuNNrHg+mRbaYw47uzg
zNCxC=OXD`*@IX%#20-+(kGQt1>vR?_{t)Q-L@YIbmZ>j>j&NBAm<nB+!`=g2x;(9R
zru^R<1iU|x0I-ZR;V0->Iz3rd?gS(Vvrd}{`pmD+4_=ttsxVGohYbN98z;z2jCwY1
zf+XlZaNMa5(s-BKok;}Ce6ipI+Mk5q#C4<nNovr(ozVk9f)WR<hZXeo7)#zQ{Ub>Y
z1p9-J=6ES1sBuq~p_E6cn5PM^C(Xk;!3>9;R7J#*^cL!$EwxoOw;lpx$|O^LkzeR1
zFSLp*8-9Sv?{1x!*+O)9oj!iFhR3)xh2{tDb&U_Jb7p7^`*DbsZ!14YJsnxtezJj#
znhiZZ=tgSA#6BLD7#4d%duec!%Upp7LHPE=Hnxceky~y6Dj;S=O_;mI%-oL;Cr!0*
za;}Ye=K2$mtc?6A=6ZrNg;0Pi!7k#8{cs*iPvE&6LhQ?l;?OsSc<=|Y_B^f66;wh`
z13=2lY_E(*TU+Y=B+5J?lQT64mAz^NjY<YM=q6Xj-?m1Pi`MDofYBT<$XX;_94gy0
zar7%M{|;6~VW+e}uTNUibM#Tc2cT-Ia(QRs4q1hZYdtZ)Band$*2Oi_VbRyI9mq%v
z_9uV@?QDYTz?orYVg45$2<WLU<5_qGpZy*^FPir{8=;9bsK2r$haW601+fUdTXII(
z`(xoi0DgFt5+vPeW<yPwhg8twF2b~&e^e+rFxR|t_vqJ2G+heqyjrWYL2|f2$aXy?
z?_~5T)Q?h(`YhrwAw`VX;L5@i`>q$<KD$vx40};o_%@hKBClt2hZM2ns9t7N$g&R(
zr8UScUon0#^UoNwu(=TMY|H5zVGC0QcMn;eR<)soSyvt=%3wnr2`<SYtA@%iD*cD_
z9Q?n7Ip-nc=MSa=b;ds$I#QyGn#BJhFR=MY_Y;@r7dVM{7V@8rhrK(yOGZqW6rAAX
zsey)g9#~e&=7o7EwrDZ2%v59oE_MDm7pcwaVt7f{;ezsC9MsLykVsPlD}IrdL5P$v
z63PTJeT$K=Rtt_o<&fD)-tWCA#L3_uDHim}cyC$Tu5JsRv<$pZ*fhc630XBu`g7uR
z${2VlXLDv<*|O8w<M-q5mmK*KA7Zy-?w?)JIz@wfU<NP(U}kVTC+7XJ4EWO_aeJBy
zJV*y}>6Gkej6%TL$HOGf-qv27L**uA)D(g^$jo@V>&Ww}P#4NQJ$XF9Dh+nDyh$pF
z#N`0ahZ$VEUj4foD#i|n<(g_O-058GRJ<ZbMXEnO9@T2O{y`?vtEGLERHh5oE2MwL
zm`PKrL`X`#?80qMOn*u$X8q+0b;`N@c}^Qi?IR(_-Ur16U-~~bNq1o~8vm5s-9diS
z<AAt)GP)ja%CH^xxdA<LTwd5^Ug!O+eKRCWc&bs;QqXl;N7MGi9s#q}{KL3M^Ww&o
z*_!#sf{`AlRSw7W0s*p=e)Q-w9qJb*TRS2%AE}2BvVR{iIl30~<R&jyrW09h$?q!m
z6ORvujF#;q1u|zD=h&{x%|)Ry_*LXl^bS5&B@}fiuO^xPGd9+iY|#Xf>o7M!^<b^X
zI0YM|{qUGCT)21{BfbIeEFF+8-QdQN+)s$#%9z6|WGNZvqwYpbdRnfcT{bK#`q&*#
zD^@zo8svfJ)6O?J<ST&GjTepp&Tg|&>87wDJ@<ri!OfZA9EEiy=t=G~eD(8Q<c|>|
zKyH>xhYgPaB?BN%zz6ldJtHH6M`RSdtS$I+!pEH9o`--Wv20zY#bMNwCfu{$Vg?!y
zM-b$0UZjR(bylsCoz(pN_J{Kl7>1#}uQ_-kCDAEW6+i!=YASi0*{xkbx<%ve6o^at
z?ogfVl~d(!%Ap>o#QBtbnPvS^6Jw3<Kd-n=S#jq`YsRku_6_zli=?{LJz#4CJ}8+)
z*(oN4v~BW)&=znnbdwer?|wspU0RwR>A``7@@~F#-yoRwD#>9Pu)^a~Pl=TSSRB)d
zWhHv-@^<yUMc|l%ie1+XX0uX3Mo3yqYbTwi*R{N9zA{)V?|5914Wy2#&n^91*1zFd
z1Znb^rHrjaajk!y&$8?`!}B<?7b`4;tw+tj`0{zA^3vCM^@VPfgrTwwf`D1Un||@k
zqOGm6giX7hM`yEc4)2vRxKVlV;J_i^<tIi!QGW`!>2+w!ARR)(DGUMRzV+VeWLI^r
zf&5RvX$+qDebH%lYuD)JZo!;Z&!!N6wM-i#GRQ3ruqp3vwbeIP1&N2HZD;^A!%|Ki
z!d`t63k10j6>bGQ*R2NokH@<~B;Cnhg-{%?pf;!IT}v;!Aqbj(Zs$^(%%7WE|NY-L
z_J&6Ff>=^b(r{&`KTG?ie&u%td5(r>E?2m`Z~Mh5Ek4<8rF}rEISxAt9q6{Rh#t+O
zjAG6$=%OA+J;DbF{l~_bmWfSYNx<|5H#LSUgP53_utn^=5dE9s^i_blIk&1Pi(3L9
z7!XYC{Um`j&RzfR{C6n^Ql^KupPPRT9*n2s=K;`>01nG3+w)yTnrhr1-<3!!Iu@J|
z@2u{_RW9$T&s)Kukx_IxDD52-ZQugjUdPQ3k+qWFW-pH3Pg#S{@7_GeG0zIG`@;(E
z+OzFnAU189xm?&ZTmmGXY>qzBQ9cIF3-A=o%Rre)Ul_=D4<20y2IIg7wu(uRCO-<M
z=|^>!V!EX!O0SWJELABeyjxaP&yg7Q4F)U~B;JSyBmoOI5TYct3*;}63s!rg+ME^@
zZBa`)LBO*0Wuh(-<Fgi%!*qPNbJrw~j~Sc$E`qSQ619RfCZ7D-pO{lvO>Zd3NqIcu
zalAOdT=~hs=&$324KFILo1HDn@F~4sM3hCUKbb(m-MBepxw{TxVQ)SS1Q(Se_Wc6#
zFG6s?6y?uD3<{31`Ckls01mndO}#Yi{zy#dl$Yy`fr-?pmQQ$}|Kli^!FPa@mQ+%k
zgW2uxSE;EiUPv|fS&>~c7oXbv+Z+{07a3=g<~B^x8(V|}+^;IphzKljB)?0lyK;s;
z83_xW-Yj1R03vX-^A$S3h>Kd|L4;H0s=J?`TY^8Sl=WWigdGH1TN~fzd*OX3ZF;r`
z;sf^PIK1YwJ!V|FINGb!{(UtDPvDxmB&G?cWA5LFM?KVv=sQ1)7`?Xd9+lFVsO@DR
z?V>Nj$ruqbv8X2#$R*0NQV**lDy`-!weFX?KT%2!AmKfFc1Y+nh=3%zFtZ6e$8r=P
zlFsQ=h=udRVjd&qCEL~>ZQ>yzAzjMpZ@`z+``4GF1J%^hN=w$exQUDx@!+l^fI`Qz
zWykG&?2!Y&ThMy}v<FO+GMUXOw1|zg#B_pQrLO)N9YW&n$A31A*@v)?;w>C2u{X!e
z_7eKBqt__L?Fg6_(pYfwtw_1JZS&sms#2g8!3XO}9Vsc`e|~L7M8Klf&I61llWm_w
z9b_OJ`jm}QP1`=gKkm}oE!N1zN<QGKn5j}*ig@1$bGL5cpKVs54`QtV$r0%2=*p=j
zum1bDyViPa{^{Lswqu86$(;~VlwO}#ss8t=Hos)Q^p0n&lk+_FquLMNJCMK0Rp=7t
zJtAPv$oPc&2H;K_EW&vl%#%iqCI`YfZpLq~vwoU1q+USr?@~VDW=OSe_908MGdn5x
z6!wC^<r*;PX}MIfpREcxjR7uv1IWMwJWz1nKL>g=*Tu#41b{_33R_tL+?{N?z}Uqd
zEL&A?-L2nle9>w?Rf4s(wXLnK<qv2BdZ44}{XFDCi6r>0R8TbQ&%O_H%U!y&zcnG`
z=0I|-ti^$Z4e;^zOD_KH$;gnnYJ&^Qxew*H-k2ZrA_;ovF(_!|{~=pC7moIbG2RU~
z>t-LVL+&9bCvJI_p_r80Bht67ZElU7ELX#4Ez93uc;qoC=@<(84PkSg5^n9g(2HXh
z%g#?ZAWkU|bySGN(fDA36AtiCqlgthpJ1nE8Uq{U9YL>j4$sv%<xfS*EE9zBu{iu`
zD?GM^rn+HK#}zIk`cuh~FjZ;O6L!<hVq0`o&<rnwEfv7Kk)XqLs!63SFo#;;mqGk8
zowcsBS4;&vEVc1{u?;mpbD^fk-t2u~2s;RI0B6P@hK#KBi2lXx?UfewkX?(GO4<N{
zZfQN$ef~Poya;H3pKHfI5!<GtV8d<V(?0)6GMt^g!UT~Y<uPI-V#AJ2OdlfSF1$U!
zr5V)zDP?Ge_Tz2rYMiBH6)@VyxO+xfI`g(|2X{(2>CaEo&yj@ao~#5-0+NsOG)|6j
zab~G|SjW|oF5ib7j7qoT5JAD`oLZxeqS`jBb`YdREtwwp!mL>CaL)ivtJI{DeL1#u
zy({1+;AsMPTqAlO@Uwl*l)*(-d0X}R!-WL@d-I3`4wVpEL$9S6@>~yAg4<1|Tie`?
z;<a||e7aUA^(;bLaapi9tJe?9q%3NZEh+z|!7`&oh3kCz-t_JAHx;6y+c}QC!1-@y
zSOmh)l&_bBrNO0vrr!^OSGXEP0kdSxsjB+OnYi>0xc}O7*iX)rx4heO6pKurq#B<n
z(Fb9yxSUVH=jQO&=Y|hHnlB#3KBg_V@hxYx!cGLx<1UJLpWFUjoOR{2dO`GmzP30P
zD%I21a}?w;Miar=m{vpW{tq@0zq=!aK38uMReCC*YQ;2(brAF5$*HxLW#_a@5WZM_
z7&eI(U6%{SX%9Iu!9I;9KKZCLv1VC4T6)H(ug?p)i7b8&ZXnS5M&V;b!t4g|Aa57l
z+e9y{N0IcUdo|WivajhV8MO!_VdUKc^rH09l^FdCvdNzwNZXTJsx~brEE3YP{1BHk
zoioFjl>UjJ?1z5>)>vX?gx}c|T#}Q4fg0NO8~h*dm3*Z_vK7&&PZn}1UG(wfLum;t
zQ9b18_oCLws4yokU8|r-h6M#&t8qAbEFSju+~JKNApTuf>p36|%>LZ*m;%nunXDc+
zuW@i9wgg!0+pO$@PATL-u32zL_<ado{;P!ILN(ReHl4K)QMFDtHKr=2pvSLOTKMkJ
z-ydZx^4+0%Sa-D5%$nHstG)~XVT?_u>A(Kjn}VQ{h?V0K>o)YYC!@&co>_ix=?Sdn
zlW4wDN71v(pIlZVu|O>0Kws>H0F(Av{=(j%7}laS?lJhdop;!bU?loZi*Mv0#+YjA
zl@WV3G$#uRTCSaQV%4DA4NJ0-J3S6%j)^hKsAR%Iamr=f3<`=z6xKH%TV5c=kaHD+
zZP4iqD>%=Y5Q6y26V5+G3=HS!1XtV&l(@_pIxyF?y38xLtb@s_&&Lnd5o)KIG*XKx
z2^g6t2}#Fa)^-^Azo@o`Q5KXJF*j^<TitcfA3+vKo?NzOBJ!O?WFaRR@-^hD^cmh{
zH$-5NJr#R3!E6LV1fOhJ4=E214|he7WiLTMcS3Zp$kvjDWp3@H7KOwXp<p)(BGs?&
zL)$pMSNTMYoq|9HK_xzikeAU}g93$vLgTE6;U+`|+nulf2U0oGAtg7+o}QAzxziYM
zaiueWyE3^kDpGQj?B$L>5^jM0S8l~(=9(DHFQSleoG9>3jx?P@$6OMgis~YBWi<p|
z$<L=S44D+Cgu#Sz&`C|58n4IPo3T+p7A5A>_F+>s_`)8!%E9=ZC)c0*gOB{d^H9YT
zAt(Q|JaiP>=r3UjDHvl!?1!ED3#C|a%0@UBS4B1Q-pld<^N5DQkMYg1i5E%+_3=|K
z6`L$x=4F3$%gr$6eNL88(b3SD*x0H%(eIlcgG9+p?OA-j;|MPxq~<J!yCfy)Dd0rn
zE?>3?GxyECdDe6bB8tis6U^#KAOu9(Weci29*W~bS@B+1FUcTeH-!wF(8)5RwS%Sj
zY#HtSc~63h?{hpl_PbE$NlHI%*n|uC_&|>@paDL<gcaQ2nmoA9d^%f{`GaTYal^MU
zKD@y(VPg*R3M>$L*0tpFSUsbx>#~#OfoI5rO0;ZjXiU8r1cHN)lL+8(jKSBwv4!F>
z%PS;Xy`i8ehfN<v?NoG1NSp1dS#`^rUcp91uzS8hrAU`62;kl3*HzR}CJj;%IH=*h
zMOO0!uJHGH$gzgUlJZK(U!b&#EH5(3hI*c9jr!$^B%rrB@|Rs)<s>A(EqS|?R)z=9
z3=Iu!Z*LC{4rbGrYb<~L{l`gPX*N%3ZEbCNd3kT|v^LrO=QO#7yAhUqnR4|*L{!0~
z_Z4)NmCl@KLG0s~qU(uYs_zD^B>)UyVGCxE4huDpbQY?iXqs{HHmzr7FT$S~@x0af
zNjY?W82?>XlSRwmS($nm<^6KUq2(np0g&I3-7mG)6rh!~il>*{@Ta+i%D2b3H}x+z
zd|`58E-&@%H4)PfLlfU6(LJt}PSq;16@<J_Y~5r0(Aeoq?UnBFx?m?ep#3-X&=z9(
zpJ#e{IZROhL4x@D?+jJCxyk1yral~;Fy-Wz`WX|I2{1}1-s!hDU1L@!5oAKVHjbnL
z?PJWK7vf4%WKc<W4t{t)wbg<JxQIv}h|KT-+(=7f1pU}4dGz@8Mf8<RA3*_={wk`W
zh3!Xnn?ydqouSge+{NH6FC<M^m&_~pATmXN{5(N26YUW*gGy=g)1>F}YwUyMCoa1{
zyI^`T_`nZl*tJ`s1pkg(up9dKe;z+eyOEdewN~8Dr+7wRYD063R<={93A<{DdVC?_
zNdS5bD(mqe${_{$=eW;iltUUT%S3?+I2qIp<v+P8*zk|18GBi^%EA+xxOm!ryy`w0
z>i&T%m!5TP5}i4TqMksgi$~>Y_WpgXvU{tXU^~F+^4?c5YPnpq7K-UG!kB$rC4_Z?
z&8$tWsgV4Nii-y`#3C6S6%-VlosS37gqj-9y;l_{mFDN?Z*OloIXQ`miG4@Hh&Yy;
z58f5X{Bzcb8_M@c<Lp5;w9R3XbmvESDs^P{vCtQ)KMNJ_ru7I)vEeco`3SmpJ7h%E
z{p_yT6}&ib)`YFr#iFY1qThX%?nud^O;*4AiY_qt3qJ}u1&S@lJMTg(wUficK3<K#
zQ3#gpXXOV+4p3zUG)nzpi%Dogv|zQ#ij|1cLqM4zg!047hzKnBi^TOi<d8~tCYf8G
zrJ19f?Iu86Bs4>b=ou6j=)^su4CbK_cw9Q^lNL8I0R<n95hSFdI=7Ig05LeL`n>y4
zpxmob6*SZi_;6Pif<cwgjJlL}%6J8f1t=E&h*)QjP+8=n`SKW!r)YbbwX(<7a4j07
z;DamG&;1C{Yi~w1jt?cWL=%GwUQ{PYQZggW5)csNt7NcqaEN&vTgc0YbXHXvJz98o
zHM`HN;(N38?8NtFjFC?3zV!z|$5m)I>SRmw8O8sOpIm~&^-{FS<ax<&>wVFVvCLcG
zony^1=yX=?QV|{;T#Jr`{vvK(Ao#d2rCx+{L7?~t55fJYS9!d}vt2adu34g^#^d8X
zUTMh|3Bw@N1di-gmQF6PSC5D=>RpjYR;=smIO>eDiR>y)!hYOgfP5K1kQx9D3od3y
zIXIZ3R>LZL)DILRbK74xFZZceu}YpCSq+Cmb%!~_1@(ETV1MQmhWS|hTT1atLSszh
z3OKa3lvBqCrf6<#bBEY;$~E$QW8<UC%VXaYRZ)o=`=T!*xgq{0-sRY9(Y@Y9f4!iz
zd`>%F7dFA0>oi?L?XH5e?CzgoqLs5#ye1*YyLqm=yMs%r2`z#jfgP~}PR1CykR%E{
z*1<!-nL6Iq8*Xe;cY>2}(<y1+sKy4pc(q7)2o-a1P5Rx<w0rq@3Y;7}ZS<Cx%F3#7
zz1o>4fY#kDPDxEPd8|{mCz@UFffdlyWco|}Frs{P?Yfzl@wT21JOf4c(w|gA!=nTo
zO3@2oAS~phmR6_Q#+Y1I&&L-P)mwc^M(VsVlm*VhAC)NjpX%OAJ>S&bws|#dd%wmB
zEO`TUlwwHg`j0Kch9cg#1%K?hASDL6Ro-6lH7fDzfmYF<gxt#=aKT%kJ}cJVDe=NC
z(b4v7p{AkP?eY->wUH$Zln9WXh#OGTOeRu-aFml+@LZ+9sdp@(3*3>Gg9byZq%RWF
z#te$aUT9i{MHa(FEIXN3g<WMdbUPD8-m8Cb#uxw0CN`&K|M8ucw+tdKvq08>2qw<f
zb#|o+nO**X<=f3J<wp$s+;gT**f^&HYDh_=VdDjT(HifmOP+>~y;x*q@;rQa@&1~1
zDlB^u|7{5SYh@WqB5jMJ)rK@%!=^yz!f$r2iaZHsT3UP+KJF=Ck^0xg0h$?)iM=mE
zsA#r3WPB5Yah)W7dU~U-v?FjjF-4r}{;!zwv@xNf=(lJiCQII4Rf;|u3$?(~3ItYI
zGFiH#SYg5y<!1?7wz><pJ`9h~AQmA)iW-}Wg+<CxAs3e~Y);CiTUzv3OHWT%F0q%Y
z->xbA+*%ppp^j2fq~Aq07sG{yhKHL%$7DKjyOWBfGkGG0pH^U7kUfSgC{sLn*t~4B
z|C;v}?lQy0%Btj1{dvkzbUNyvJK+5EH>v1ZyU7Ud@j0VfEp^Q=G&PqXtvKd0kl<l3
zgNj1-sinS}h%_!PZcI*YE*Lv0Qj}NkPucu8kihU!*#ZijD~CFBpPS$gMAI0k;U&g*
z7LNJ*P*WF4N){s`7JGRWw~4{<ebMcaub;hl(`xb7^0`(!S&7s#^42{Wmerc82UD)z
z+R>o4=%LgaB97*~i}spLjn?h0=UOlx?Ohzx#IPt3WwX}h+c{q(V~wZq;?>O5XrrQP
z@6SDtaM@6;f~m5_zbMvBZEK~Ya}X3^NN|*A$}ujO55OQTHNnC##DJ_3CW!LEzbyLu
zB%uAQw*A`hnk(rel`2R{qBQ@DhP&MgtyHdB*SK>yy_U3j_3KaL3f=oWu%y2~9x}N1
zFWJ&<Z|;M<q$<cT1%CG1*pt-Yd^B?^dAy<EDq3A976-HXP{<>O<X?jm;_K`$Zt8!$
zf8KHqnu-&{GL>F-TU~VKn#pf()?_2%6650&6B3As5{U@niSXmr*W%Y#<IGLt&CL?>
z^Aq#)5*Qc~8R!#I`qyOM)V(w+9gVl<UbAwkQ`corO#ErG_E5e~$ABF_0Rw3b4UP6m
zXiiR4xus=?yv&ce56e!mT)h<)AxE<Yj&axT9c$YqjQJ}&CAL)6C73@jXB8DZ+1XH2
zA0(L**w{*xxp@<QO22)x#l`OI(m(5^d2w+OJ`4Pf?iW%6nq){L_i;OFc-^Wo7nA81
zbJdtGixj^+MQNx<*Vd=)B=2`8L`2e=M6!V##l-WkBq<s1HZJU5Y*=dE94Um48K|~D
z-IzwYy>M%ei74_hziy_puKvY`u(WK{X#nMAdZyASMY6YyyQkth-3Xo3eTSoL5q<Kq
zj06q>o<6gD7Ja(md(%M>Ra0|zakPF>ya2x3Ish#S+~*EX-AB*_eBPzF40rIQNMXab
zVG+_8Ft>RJfi&?&vg-bF2{edT?`}~``eY!34}VEcP~8Zy$S+08tip3%%_vk?07eC_
zrzep=S>F=DU5Y^t&y0KE4&gSdeI!;$YDI5X^K*inVGEL^mph^LZiUQpYM0%NBc|KH
zupIsbi*26J`DQ#68mF%#=oSl{JRk;x`d6eRmV9`mS}S0^_s?TwgOvQ{=c$OWC3^@*
zstYvoGk38mKnvEpO2<dQTb0GAf&t!uoJq)ioAI1!sxva2aCzzI^xTWZnw}0o63~U$
zx|fG%zl+4h`A#D|-&Yo_TiFx{2eA>3u>vVM2*({%!X3S}TcoHV@Wwv8Z)nQuiBgrI
zRoc>PBK#7;1!UrPe9y<lKeL+mMcB`c=|S;J`))i!F2U~GnXtCSY~P@vS(%8x+xQ!|
zcgz3th`S^+V@A7Vq)NTJJZlk8f9cn-eXl0YR(V-~RX3gFt#5w$?(W<Zl>r)#d|ezL
z(!+ya+U;)7udDbww#V~@`@vEgB1D`Zu?W*W|5mrSzC;Xd0c=!mfxR2BEhGxOH)E5r
z25~ybd5X`C2`dB%v-*%(*r#LfQ%!(?u?FX(?9*_J^Q9b-XD=>G%XIZe>6>2quqnZK
z<1-4OQG{Xw%u8U`;dTy^CXXle(#%cjb8JrEAVHeAq%Z4ds;05Lj8ptzzzZDS{so9}
zGl&`Tih1Eam;E!ZuTZ}+(bNA99Zh_ObiZ43%ma0?q9!5;))0$qwEt<jRrsXDZX;&)
zREVkllg;KkEJ8{`AV?^GjjzZ~cURkY6@&-O2xPw6M7?u!JFjzEKQFTjXD%*sQczYr
z3-xr+Tvcj%qq_RL*#Sgm3ybRc%N<^zEd#aCCxKt1D^jnd4d|ps@V&{OJe=~}i|BOe
zx>N1!>`W8&x+F>QeJjyQmdmsycrg?{A>U_9fqU02x8(Gw1Ed+9{u#nrFivgk7dG`Y
zPWfsF$entht?*Jde5NYMmu#-b4QIR}mH|u!+-MlDThpK5aoYAw@VMl|e3O5fbm4Z#
z|9x3dzKhV{1PA;Om+|Nm9hEjkYTVcF_$wum&S5jxyUH~j)hd_l_R|~<W?mJeyJY)#
zu*|dfGE%ue2Mw(<VyV9f=^RA4#NPm~dMETLN}%~3k+T~G=_cw7l0>#OY%UM+@YF!Y
zuasv(`5S!*jYmJ0hpF?g-J8#rI)gi*uGSf!l(XAglPX?ZP=6{N6l-|5&(6WHEp;?m
zPv?BQOcx+v{_9UutuCKlApwegCs~WNYI~F9D~K6ygSCCT&CQA>pR`pZPWTR=U6=4X
zyFrg#7tp|mt00*_lBCoo{HMBO_!MVT4doW~Vk9yv8wtPLTAq4DG%(~K-kR<%;o~JQ
z^P-OKL{3yx?&uMQast&eqc2|aDNcVb<hf=SfME9rCI*Ek7iXO>NT6>55pX`+>`IR)
zD17SVOL33l<KmF^=3CrK_Tx;weFm_vy*&<OV^14Jyp8zGTuQDwnle<TZssZtC5db@
zs@vbSS?z0)zXWqU<Ss5gzKZouRZCNou_{k1<=scU^4;e(h8*FTK5q77B8ZP0c;D8X
z3OK*J+gO+J_u-{8`Sp3%Y1V6E!ulHCDkd*4ue4L%{qMWX)FRD)H8EVWLKXGF1UW<u
zxqoAs$E3-_=%P-8`&&_c(dVM_@}i=(g}c(MBLXe465iVepX={5{=?J46GRa`($1=}
zz&eMENFvc8i^VGM!-o!I2bv=;Z7nS_0Y@p?hwe6qFL&$QbAu+!tMm!WUjMduK547l
zJb*FTHPJYcX-^kPF0zGro7G;TsG_paOk<JTMfmD*`=xI2dj|G7LAb6+O7jR2k@>CF
zzUel*=VYsxrKP1(Ms4c<TEM>?OI+NLLgf;D8b-08_@5rXyEH}=v6(J@A@f^=-|q{O
zm~F8DPyVM!SkfmwEr0~fO=KMVU6~IkUyTb?>Dfq2Bj3&0$OTUhri+BN3bL~H{VYyM
z`;?qJ*YGYsC#Q1pJmV+YyNZGWMrLNKBeQ0uH#aC|SicAeNJ#W_bbg)t1c80=f3BAx
z*~}r$9zLNw2wWlt$z1RuBo{wzQ&__V-1%AdnAT8ZM+CC5RS{5mIIKBp>-I!XmyBFp
zU00r`sZ!wzyZ<p!S64SN$;rrA$Wscha(Ozny1N_BV#b>iNhUP$<<+GgKvP_{#tJJd
zD}#ffPoSJUhE!BkwH^%0sjcy=WtO7=fdAi7&5S2yFrOF>r&294f&D7CbFD{n#3EuY
z8*Nx1&{WfH`u*xwl15%~2$$oHBQN;NtW1Hux8!=GqXp#8*_oL4wcF?n9sA41;9vDQ
zIY^vMGTC?`QVxnIw75_ct|YyZfu5eWPE^5{4qT?@!RprD-pvwOP#hc_boA+FFQGS+
zD9Gw_)iyJ?H&?mpOnW4^Or;Up#Q(pCwX&|Vp2679T$9@DHZbKSW}p<%#qOL^(hPig
zMI;~oMqw~tt!h>9QB>5<2LxkaN^lJ(f@j}p2&a|&ACL(;Z#X(paZ_Il(d5atRhSQ^
z3poAe)-YU3P8E8GJ(<tcZ$0orON)RRk&qdTnXWsZI*pmE=uT!t_dUx(BtH<EudN-(
zvuD%DMl3F;@;g8#xAgW%{#OZ#upvDlA8h*oIu@;^tJOr8wGpjmJiC$qBiK)RD&Vby
zSKfpwOkx7uIhlw0mmQcdAAc>~`?|lnx=N^<7F$>XJ)hRQ<h1wTE6ef;>=W`@V$E3P
zS}q&l0prVGBg|jlZ$Jw$S)tkPn0l`9fouN>b7R)Kw;;I_gcit%D^BX7^0)&k!R9i9
z*dSDIQ_Y;~1DTL3cQFy^qtU93pw>+=Y21D@pHB)k;}fnX05{TKw-QJ<L++B_*5m9-
zvyDk(EuF8<+M0<hs&D*-UtDUAO-xdrSXAF^a56PB-s@I+8TSU6gjb{KnboYqmb*o|
zRt9U`*LPhZ87(67NBK@vGsVg}q+I+{A5R7fF2H{3JIne(ooGtT{bs{yp%xhM69(mc
zC3gVpkn?76i+X4nwTsO$?CCa|g?*>h`GOX_>2Ytuv3<YV8Gj6kKXNZESm`RBEG;47
z{WASC3<GoZU~(=kMJywQE8vM}>h@g9)rI-q(O})X`K6cc0iIh-d(Bt1GqyCYcQ>=K
zDvoJ_4t6d5UZ=!oE0MEeclMHj_K)c8oqpT@Omup}Yd=%1wLhm%MrPP#t-aPoq22Yv
zQR(J}qi~*>$99rr)<{8FMC5(nCMt!9SJxcF{En-~VQ8=Mt8o_%WAG4~Igz#w?wdXq
zQK4T#7o}}}PdvIr`3*qmn_7nLYo`uQ0WAQy1Bp;%B>f}DMYW@;xTuTbCNeXGd2k{1
z#4_kU!W{+rC(68g<(zquH10-6h!sxc*5;@H-1pgI{;a5(7K)8_cQ<<sLi<>HhYzSu
zEq4Q6hsfe-L@yV~J%Sr<T=yb`YJ(`~Q|9U&L_F+UZhl~qc861_ETST<ZT}4Lx98+c
zRp`vkbs&81Vr)GAZUKlow2T_vynQ-)Z080Syw?L}Y=u0}ni$eO_Es086S*qtwvU_5
zfd2&-1lcq$m8@LXXOgm<_9w{o0Hc{1Djg@Tv18x^u`mgBK;~#MmxG+ICx~(yHSFru
z{_vDEygfWU9q8?K3~0MlBm-uqpY-m1;|hTV-+;q%z3az8Cus!uUqh(Zl?{)2awrT2
z2?C!~DN}7rT$+>FkJM)>Vm^lh4eayYR{@^;B4@21fSKIod6|!4@FUHP25gqTPupo~
zN8td5)5^^J{AS$@*@A+Y@T%<PH>l#Bkm8V6sqrHH_f!AA)g9v>5{i?IQG?``r(arW
zTL%%p%g55BKV8faGe>1NZ0LmhP=c<H;87bVr(xJFH0eyMHBL><dhmt8V(fI<KSdj?
zsQWlc!1|O-%sWjFZa@fUH>}4c@LFCWCnERW8EHCuh*ZDd92P0!x-@l{=KgK;@ST_-
zkEgilYrKppU9NGV?BDvJ=z}_@t@&;8LD~1N!ApOI<`iO&eculcvcNDmC&0KA!uc7{
zF+dx3TcOPNKFk3~pk^0WFB(2^TE~PR^4tI~nnU3=-~<9}SoSDhV2EWXp(RzmEpAx<
z?QveAMv0p$>XqyQe^Wou%g!+%{6f>d{7cb|_t`)H4HsOfT3HhhF+75-_+?G0$8SZ5
zn6e}#J5ym2)TH#COt^lm(CzqWZsYZ+6H%JST6k!FM&0qk-5==j`3jNI`@{ApcB!7b
z;VT?X-q#rj{~up(9T#QSg^em5Qc}`glG5GXAUS{_A|=ux?GPf}B0UJAbc=Kf4Bg!f
z-2>8b?$P&o-{+k3{l57RX6}8jz4qE`U3;x-?@bIX&?vpp$H0I&D-P7gL`xSa!CB<X
z4r{z#zGp+@==|pCK*cUaO~9VILw)gUd_n6`WMhHWqg&<&VF2h1oLUY6_VC|{sUNjE
zTlc3~0arj=486YqZgOtcLB4U%{%NyXS<*M@Edqr81Z74bbXnYVbiy9}<Gesf;KEqS
z0#L&bpGgi)#?>fA7W!eKrJu}5X80d9{qA=!Q(hGKt^Ddmr!H55L^?Hj9N(S~wOlVk
zrxry&h{vNB4P~rE)|CHoS0s19;UKieGOJoLeD&{ejLQ}c0}hC?XDdQ~p7f{kJx;aU
zx4sAB@a-nLpQs2du%iQDiM;Nh0|b9<k=1aeBqb)kx(1~SCPDdGa%58PsTO>+w!dGZ
zTcDXK`s*Sr9I674LjbM=OuYsJDLGlc#P~UmJ3(aTM`<V&1&m9zT3C32Y%Vi}2`=pI
z9sKE=;kOwa;pk-8bY<_4Qh2c(9KqfE^Jv}hTSK7xX)d>B8OYEW)J5v3@(iWk;FtBA
z0E`SY6fFMk%;tAgADn7{i+p@19Rv@jD_$+2mA!t`w9()@WW3JjAA0!x*W3qg%dTQF
zLC5V$J9a;t($mb%gC>XlPaeBM0-cib%h{cs@_y$}?s=YxXgva(0*t*s^S9Tfy?)XP
z<tbYMsHxuPz*Ljog$V^=+g<xl3ATcieZJU~RUNvQ-M|tHr_7z6fSYeBS2wfSea^(s
z$GE#20$q!3%CMeWCTit{BO!@o=$w=Q+$N*6(>bc=U}-v)AihW7ot^ej?ANbfV})8?
zpY5j$9vA)gb)1H^MH#ylw@p%`KMcszwNp`n{)!|eL&&H9k9;IDw2@zX7n*$T{5I*e
zRaGaZ%p1s&^r<9#x?Zh?Yx$VoT+Qv2HQQ5yxo%H}zRCm-o*($a$GBqxPR8maawfik
z%4>nkt=}ZK950j3u>c3XR`QcRWxvqNC#QQN;dNlDqZ4n+g@=iG3@@y5kOIOg_7MXS
zR%B$P@fsfC*l&V=-~b8ule4L;nON^;T|yx?HYO6-Gf(#$gSNYSA_0c-va@3J+yw2g
zLh>1GnZQQ!z8*f*QrL2P){DV!)$_GtjVp%QZDXh<HsIzLG*pc}GSb_8>Zu$e+;}*k
zz(<tMrX>k57`fF8txnp_6!rD>ALOic$J$s~DSsUu9v+53Xg|oQ8?x}kwzS+g`(Fc4
z+rVIQr6X)8L!@IlEPNZ-CjJ*yOweN_Wt-R7>;L|uFsJ|-*JEMOPPAz)()pcQWpj(H
z>36agDs@*#&I7q>I0TMZTkAGYle`_>o7wD+lA637g9oCSstzqnyhuDn#QRBrg_kUT
zu~za5EZsHAK@w!9lfq>t>U*{a<gR-A`UvVaHa9nUp9v_1hK62jP)hg^Ez$d>97V6T
zn5e6(PpSQPAmt&yB1KaB>q?}}{nvFd+;yBpMXXo1vjbZAJxza?zZ{T;knuew<t#L<
z*=f)4{^oNob+ylJgB(;Clf8WV;wx|laM1=<ZNRcneZ%2*SVx`T<L8NhqV%Di<CH66
z9f7`|Kbz_j;*}gWI31r}aBgx{eAP(>0bAAnSb3+53s7&$-c&6d>?;D*jcx;{z0Inl
z@Qwp^UgC|R=YQ8eL)F$Dc(-;lOrXzaV}M5k-1iKs%6C24P@@pcN*6rYvN6wgPB$}l
z=Ir~nvjc=G&YRycQPR&Xg<3ZG;F0@@y0Nem6`<GZ>QwqtD4K^m0;jPOCxdz!oBs+r
zoT1W12KwB_tEMQbYHzHSg#8i7D=kbqMR!b7Gt(O$`+^voRG-`QT`~r2U7r|Vn~cb>
z&8=gEyIJb$Bs(QR_X3&U<Po@3Ke)rdeS!CMD9+n>okS9#UWJDj<o*|F0U2`3T@W8Z
z&rFH5G-RU&=^>bM@Vm<!G$egP{meNE&qiQczhW`)d1t^4vWk@^Mo_){uM2T0e?BQc
zkzZ<4lDt3jQ}_EFT4d`sHHipy-HMR9ppusTEWYF6tDwX|qo03Y${GF-ywi`A#s;eH
z`M_12q9M4!(kYddT7#dxDwmek(rfjuFM*xZ3QC6Izfy3iL=UTeZ{fAj&<K6y{Z`kZ
zgBafkEt&x?ZZ2`}3ri-7PXnEuo#P{jgD^Z0=?8lPl&7q;7}d@>gf=1UTpOcfMSv%|
zT5g+Eg6!<Lt;DbUqh)1El-Un0?!27?p@A6_>@!Py9ZEl{37`3vF__=G>)6=Bp$Pil
zpASg=-!P`tC2X%m^rhFt(TYAzfYC&W$(q4s7lnRVqppNw#u&lYS}eQyyPU!|hGAIa
zwL9U2b@LLqHse3_pM<jH4~f8korg691}L!akJka3jfnG~c=imqoM8f>tM;micBs@E
zhxzgs@Rlb=elq+dFPt$l@dQYQ{`nLEX{4Gz6}%KB+EVdG(Sa6FLk?))TVd^#qOQ?w
ze%t5(xasKW2eE)SF>+|~2bWL&rZN~Q6rHd4mq%+|paB^(>3j~W)fB=DXoKac4)o97
zQl!)c_hGKWz1$Rh>k@9~<1tEa9(cj=y_jRfxR;-H&U@3P)K9GZOyT2VcrQk?{!TF(
z^m}_t%@|DVQx-WJZCH)9&umpH>GDyGtb-~X7IE(4-DcY5$p!s3TV4RF%>lJk%x>ww
zJuB6OmnhbZ)#}@z>De3>58{*#E}5DA{8s<Y+++K0r(<P)gthdB7TplG^D8m2iFJcZ
zrAW2?X}=_CO*$p+0>rF$LErxv(m`ex4R9>~d@BEE#F5e`6>Y%f;TLcI88rkW6LD(P
z;D3&*{rJZQp@`Dd|KZ<-QnDNV#5}L{#%HUD4J&jN=71FRpHxpF33$V0{vmMGlPC|s
zMs`1*u6v$RI4$=6PkXGppP6oLU*3g}7Vg+f7A*YZUjHT?EKp|cg?V@!ytDlyt73?c
z|4w@Keb7JN4uq1{=Mvv9S|8dn`;x%6z<pE6q}%hHD%tO6UF0Y)P%wV>tOkhj|H>lb
zMoBxmyg0m$y0N^@8)vxmdrXM<JZ}K<QvVg$*%Sql7JdRo3oF^Gf4!AWcT(BQ`TzSV
zpg%rephz(;<!(n>Mi<;}0HfkwPyB24vo-$!M?j^2a^e7fJ)dpB$Kc;+t#3~ZyVUz*
zmHD2W|4T>@Z2aUwsOl-wL*F~BR6G>d$OQ)xk0h<ve@4uH>K_;cY*gXsw*733UL6x5
zoj<y_<T-$bG(dQjlEt?9JQX2<Xt_1hMNAZ^$%xYVZdBo4$_2a!p9oiyu~O}kp(r!b
zcEng9P8KQ(Tzg-_^MEnC7w`w2<KW-VkNUkD!|t2B+EOD<{B#@sU&*`O?--2i$~SDb
zX!t!t&dZ&le11>==n;Jq4?Nc3gLC_8NE~4+z?oKmO$aDgD8EvnplQ0RU8usdeu`Km
z{3GdqCuR)>fAHEFBKN67fKgT(l=#|s{Zk_8pdPUKq2RB}H~Pn$uo=ojIAYkHyv5RV
zn*7by53>yej(Wf1y(rk5$@R&m@!}nNW`g}0vPf1#1&E92Rp*~AF@FshI7R0+(AVbS
z;$(8k{_Rp7|M}v}E-`Vh+3kM!UUfFN+5GwZIR3YCp^R^Z-wNA(hCURFp|cjVwzg)5
zT7HW8JlKM(Y!LIh+oQO%d~wqkzE#{bxp*y|cYBE;`E7eTJO}@V=c_89(suvDrc#=l
z*55SuD=4=|VF&2e?-+7zgTX^&LRevYBQTv&0b4LWiVQxAG<Jxo*)URQ!nOst-`WY3
zwi(W1+Hw)PJGZpKIqcQEcD2sE5G%ejFp#^yc0GevEsYUGi^01s^Urg37U0H#fvnmk
zm-}t4((D@;J@)BXcKzLr&5)Se)*zRphad0`THvWf>T2np2~<vPR<dg8>=F0=&bW;1
zxu3MuM~*W)&9|QCP66!I(o)x{yIw7BOP=Xo*SH|bm}cz$V2fDh*kL3(#GA1}w6L8Z
z64oU3x7hMJi+NT<ReTFmFLv(+oPt2Kr2G<RA+_H_JVh@H`%pRYfd_)reO`Pt_Vt<$
zs(mbI0<>a?TrHhrHH^fWyZt>p$E8Yoaf4*2UE#pQ(c^xdy8{GyQ|hwdlWudV3S!*o
z<4U=q^X2#EH{*<3RMf#O>fT+cB|wphv(x%!F#2WZa75+XQ4ul-68;DT(dG^1j8k00
ziFKB`jk?ahr>vG<_PaZgLiGZh3e7JE++aHCV!ahSsYyWvK~&H#ve##n%p*-&aydYD
zU2~>D_j=qPat_O*0V+x@B<b1xZJ_VF(eJV3y=lu4w_Dn<;I`>4CVC0)L*g7}%r#+8
zC2rdI(BWpVnHl#SBpHEF@^h+cX+l81e|q?J+l^3e?IoB|_YfR>V3Xyt<kJ0}V>r{R
zKFe>^4Ng=taV!y)*@)<1(rJpFEqDiJ#+PdVMx6ycF8C33eJGF6c@GNeKSCvt^z5H~
zO5!ySxot%@??95cfNYj{ZT;|UPOJzfu}#pScMK496)}SNC%cFig~!q(9@j(v&*Okt
zYPMH<opDjwb4JIW9^QgZE>dDAyT$-MoaeLg@j7hU<L()xs#Ho;8XHAo^7<r6Qt^mm
zc&5vt_vLyn&=v;ogbX%+ybLyxYJL_fZ)zi3K~|mHf|mFSn&FAZ0a9JI-{|emo5j_~
zvn$KK6xbqiXSQ%#5D0Sd<BL*9N)T#<-5_5ssw`mo`8c*<W?oycEEIxzm6XrG{DG?v
z$vZtw;>~2exnKM-->Y(I?tXYFNbdanWFZ>0TslZ1O+4)r$(EJ>Rq!{dhuBFoX0N?s
zr-tPIC`taGUZz@-C_g>P65Y=d8oF)AepMl@C9$3*(0ALFzIcH$E4(oj0zpl)+AE4`
zxrbSnOV?CtmQD33A&hE!pm3g$CjVg8h{(~+Y|vMZng1Ul^(v*A6NMx?PPQMP@0OW=
zLJflBZTfcKHDLh1ATv>;Kg2Y_Ao;E{(-8)AF3xUfY|~)g)qSyVvxP_Ot)Fc#n8cPo
z7??MJood;N`dKRX4(-2LEh#xEHs1B#T<EKm=5F2!IMFztugjtmvx81sCrDTU^Dg1e
zi5vu}xnwyMe|CTOoTjEg!4Cs)9I0pG-e)Hicte@Q4J4fSuKLo5L5V`$+5Sx}vT?b^
z04!iq-hi!#tQS>lo#TZBV<w|t{)z>Rq#~E&YTzO|l53BEjO(QD0ix#x$p6f1Q!DVG
z`XhgTY{q1tSC*+J(vc*5OQ=FzB(@{;=RpdKmGH9reP`qU3@ok4*Zm2oJla>}7VL+B
zRL5WvjJkRVDi%}X5|TGMTI1%Lc)Fmk<aK|Q?Z|}@<b261@HGB9QA@nfaFiIho@fwH
z@abRI(vyb{bi)H-Tg-K+@U!2(mP5Xtdq>44o5nvJGB}4fvwaOpPm9fzNK=MMu%%#M
zQ+r9aW?z4;Eh&-D(PkyG@7?V}rde`+HyC}LvQS?tKg|F7l)KC(7d5RMAyF{N)~hs0
ziD&$x*s$L|6yy@!TJJpf;SMva#tnjuBwLZ(;uM1KaA}Se6-RW*9`(jbKDGd(0CUrw
z4Ld{@c*8jVh(|-}lG;z}%><{$cxKx)A6abvhkqo_MMa{~HQtXF-W2iga@&ftIX@T%
zLEg@ZU!Qr$Q2(YBmM10VYaw;VWJnT04C{vqX$@pOm2VG2ftJ5-=b~$IHyj_$v(OuC
zRzF42c&<f_cgL|9g8bK8(~O)Jgy+2Eh||+#Ve}p9_Kj3)B;|U?FzBP1X_*1dlAx=V
zzCQQW%s9%xIO$}07a7?kAaGiGd#Y9i9CfvQra8-Y@kqjmwvAeY8t2uY$YWLij7kza
zL>=(Np$Xv*;KmoT>t6<}LAg0!b?a*;(o=>$XIJw(`8A&J86O_6_1}_}vs&o2mU6t<
zOBEX8^Zl{;i&O%bS>9W8LknKMj`dLg1&;%%yOG->^SZ!OhV$B%{k!8>Aa<<=oSU8B
zTxP{eSvSkS{B1vU^*btdwr;E;u;sjm0Lbeue#CH6j}m$oPxg$Z+r(&#lly#fa#TaZ
z-B-*^N5N)uM8KmOz?Tlyb5zc6F#c~*g&qV!Nz+_g*Z8^st>%mW@oVhU2|uw0#n8xA
zRkdJ-L1sQNi`@(+$Tf+CCjT05wCQ^Lz~6n7y8adE!jE9b8BL+LUnvBAcLjakO!^kJ
zoEns#d@B2G;J>mzKali@@+(uf?Tia3AjlwOq&Km8hp0qS<INz??Jm8oKY$sO(!YFm
z8F*IB^yx>|x3@0U5?P*%JXNkt&&E#jcur1-Qe!CpxYi-08B^Sn2^f6Y7lN34kz2!$
z_=~RUYjk6$0QUjILc~{K9%_EcZ%DWkU*OR8@mdcnXt3^fDvbAl5Fp$qzE#lszux3m
z`vi|Yx&4(fq`bZwU5n4$7bH5or}#JK_jEo0EU2yrDKsa)AwL`Ss*rzYv0T9dyhFTj
z@m3h&(u36T0TXy!E&W|uLkhdP?qYA@CHA54-OI?rSiC<mKgjg$zkYFQd=@_!8_2VE
z%JSBpfaZ@dD0Q=cXP7O4Mz*wu7tjzti``tb(#Bwf@mxv6xDy-^iZFsq+5XiMzu3z$
zoUqNt#GhZy%a-N-{D9Vw0tlV}h;gSBSP&TCh+DJ{*uU{9m&P9T#CxLeM+7(-na_BU
ziW~!?fULy>>%fDqsE9d;#yKpzV*f+{2k}_vcOMZ?DkI2*Hw!H3Q)~>)VYAwj3p!S)
ztj6nY_Zz;&@mJR4g>$j~N4=Na1DX^xvT^Jf0EBRjM-}~$cA$DSR$VC_9g($}YY>cj
z^<_D!UHp!FO=wA-FbJ5i&y7`??xTc$zk}~>3Q=c?v3qg?{<9D<A`nzrRQohz2z92r
z%0^qNPLbxM`S2X8q-X%}4e;a(0QfP=BeZ1;)&{3plk^;TM&R_~4!obWboLeZbH8~~
zEsagy^W!lQR%n{&6%7lB76p?H2$Yy0T0Ka9Dk_!$5bA9(Rn+7F&ry(jI6Hzwr>u<7
zz%Wt{QYeY7DG=xz`jmY{5xu4Ip8|Zcj4T!cAK8LQhEH|I-AO?Zm)FBc`)n*^WUNG3
z0v@Ll0I+-v`e$%*P{|DebvCOeBqRLOb0vh(o&N|8IK5IJj0FoYa^nGC=f~tXq0X98
zR&eYdjTXywU<PUv*Qx)1npy$r8h!lJ)K~}7f`?^WTOmrr95f8+ouK{tez)Y$w@(%V
zcNiMILQI+e)mPr>e*4I_b(fR5wE+mK=&ZD=Dsv*2UR=|Fl^yBj&oIA91zi4w4BnT6
z6BEa1)|NaH!!aZYW|9s_)f<oOsL}_=<7=p3iH`Ym8gF5K0&E<_8>T8$Z)p3b{pV+7
zUBjegDkiK%($Nx!$baVu&-R!<VDOS;t1e0Q>&>cp?vZD4qiL*;6Bu2x`}s7?_R-bd
z_`U8PspVudNo_#uceqcJAM1P4FH_>~8)GZ{d_Ktl=5_V{&s31qppgNl0v~oY*F9|4
zg+h>2ft8h$2BBGPBff&~pZRBp6&xoL?`slh#Cl>WVJx0-3U+Tj4bR%dpr<c+Su3Bt
z6PKTzm7yMvlQQ}o-gYY8@*{rkx7AxAXc4u#$)r+%(@T(fQsU`s+r>1>HOB1?jY7Z%
z#VM&ET!_%yT)omh1Zt?BeU`WcN`*FZu%wb|tvN8T!84PPOJGCR%=-q&Dg_k%WX8iH
z8D8Ig6u9QQTL)P}vkc<1t5_U0@YpC@QV-v(l~v3=VqDuXLo-wxYelMg(`g0@@20ph
zT$`O(CDFfk4w_xTBCrAs|34<JNz$|7+DOoH3_#<JZ~Nptg%?{t0n^j0O-xIJ=4dp*
z?m1xQO*j3wSNUN#$(EjZm7Y8z`Js)4w1>UQ)@v3#c~|Q4cUE-w2LzgU%#Sv#!aoy@
zpQJ`I!5X;ow^at#HMwiS;t+-<&-i<<W61sLYXdMg6XTN57V+d6H9YErcrimCk|(A0
zH8@E;pNXuwNpemx%aG5mikR^m81IGdoKMe~!)A84%1i0Z!>Ua43O-3LA8D<-%%;4Q
zNM)7Ga*uv2Cx)TQur1DC%g5Q0SCZHFZ&Xbn$L%|n*>*j`Iw>Gld*`q{fApSlM1%$w
zV_4(vmC~T1?TM4!HYCnIQAPo~zrCifSC;F54c{sTR9sLA)y>IM3&aJck=-K`l#{tx
z6il_-eDFw|Q-_&2!jodGYH`o-MfZ0Y0;}VlF7mzD6{rQ{=Gfrol1itqSVX4P3Mpn_
z0q~4Zsf*a;37hFB)73TEHzD`ILYL{G;#L+hqeGcD>H*udY;W5W*mGB{6BZS@s^f1+
zbh{5cKos6aS^>+9fi($_mQ)IC>B$^G2J8i|#}soRnH|mdt-suH%}|-OoCkQ041cR)
zy78lwyfRn@uj7tXKQ_hO3&g6HPLa2pl%lG;X&cMl<BIh_xi%sQ0+uKLQ^5iVTKu`j
z0zA19#?~(?DT;6u-z7(9MKOuS)B-jw4m@tA6+BgyeT1Mp%xn@?r62sJ2KIVzND@YV
zSp^74XHD{<AqP#FKrNX>IOr$#{iJ5hthyf<GclN}syJ{LIIXQ^{7~z`63I=#O&FJ(
zR7>stcCE7i%TDh1XYdWnuBYRVI5}j-k}xYzzWdUYL~{{nOL?<xbd+Aj*!L~k^sg`_
zE_hv#-W9W;-ke_|+b+J~<?n=;8zHfS6Cc;oSTg?HbV}01`&YM5q8fBm?YXjUppuma
zUG5D^;_r-vUwg#k>~VDr^GtM7Z?Lv)5X6@g;#bR(Q>v5}WpIrW4f~X)6u(N3iTIrB
zWu(4Q2tA)pg!8c&e}{$j?5*PQ*0OzvpOZ`uQA&KwKBU8W+g>f_usQ$w!CCCCUsy=7
zwVTDq*6_;)tV6=ATWO}#xRU0k;6;n7XSA1)&xhirVMLw@w|mT#?85KCuUsNdsho#j
zJcI7j4?HZvOMQo3tvE<&C2Sst)-CUB|1**#%>tv8vh=J8`G$YFK6H#gSkCq~Xdf#L
zDb~slZ-$<Z-9Hh!+i1)CT+tS+7@aY?=wbP=ao&L|0^@1`3*#{n5Q$MZ@t<msL6Ii<
z8nPlmsAVc95krNXXnzOx#By{(05ge$x#hK!m*1R5jIxTF5KRj$z0;GPH6m|AEmfb?
zo=g&Ebc6f28*}O|yT%-ruMT4*sCB>E-z8+%e-6FK-h)YSve~2yFTA9aa(U*Xq2*Z4
zyKDFE#uwfDxmGXIC`tt3$=RcX4x|mVLUhs``o0N9Q&rU~WibkIT83KVo?vThxGCdz
z+OTwW>y~mT_w@PwOU_~SwdF@94N!Yx@5xTePYk@Huyn1$kBzppfz^rGmDj)Dpx1S)
z$EYb+ant#|znSRbZ+n)J`|1%RCDZgxbg($ZaG*ZkrMB9gHA9O<EaTLKR0&Ltsj9^k
zSA7`|H-DQ`qF|4y?8)cS=C-V)z!P4UbuMhS0|pyL7|2uJ@nr8oUDu$4q<#I?dns$n
zuAmEJOuJX-xSCiF%_XmxPb$HwXkJ~nnJu|%ZU*1cY*2%;>c+^qUg9Q69T>jGGtQ%%
zn(5)5;l~%hkP@r%!fEysj&dfJ8MDNxo9d^edG|;RW2)cKb#zNjjU?Xiw>qa+lC(2~
zmPMWSUA(yZF^O$v0$a%1BVR1DwjA3N4Yu0sRJ^474n?eMWtM3QCb!nWeX%>94yie)
zEKb46{x%z?o44H90?EdO9Lim0(d!sNFUC6Pye(5YQLtf8(g>chR1}FV5rAL1(EbCF
zqS6l1G)ca@LoQR@*p~(DoD?!9?N@WxdLgK)aq-Q~Dv%!m@b95FEWS2t=bF5U^nEq3
zyrA(&<(UU;Nj&$Szxh%RDSDQu)S34doJ?VK*v7d`9`u!F8h1l&jjv;{Wd{S_rCvYA
zKFu@ejD?flHMm2)#9?S+-mSgGUqAZ_XGM;pfXY@<Oo!iYzb~(m=~S;|p?%E7!tTMS
zuPQo&iT6$Fa&hGo8-d_ujg~?{Kklu+P+hm|#x@j}xQ>>I*iyHj`jg_)umUpdCEw+z
zFuyUigYtOv5*WIp_3<sKn98RFM^EN&#ISD;P8L*wXhk-H`z<pJm;}O68hZprIZQ1U
zoamoN!#H7d?0ljp@hqg_aVKIm>fMz5tbSnn3Cl`VrGqM7IXPH^BX5W7B3}aNnaJ1W
z&3i+1w=KxMiG#W!&yYOXwzW+b7v24?73GH1exCFD)3f-mqVTsrPK=4%beAn;H|NR1
z25CapEEvO*VJr?q+Nes%S1a{b#MgCf!Xu`JrsCQr<LbD3R78?p2UMnv+DN{nOq*rT
zK6d;<&J2!!AGt$-?WS?ZKA>kt$}{i4@cdc>udYuqt2|DuM4kAemBJx!GzHmEzy`eh
z)PA3(&!fM&W5Yyoqj$yiN5N(ArT8yXwQgb<-d>5q7A035%K&V1FJvXeah&J=IfPMN
zrn&pm%|IJv;75)LUUU>slzgn<_SD<bFp+U>3ih@0H@z5~!{34E7=`mMSMt^UlLU3e
zqZf(v<W=gp*fdlKKvbFV=#e@u)l}Tl_1*Nn`3vjq8!?tD&CFNyEE9QV?cemFJ@kYv
zcfS3q33_>&iV^EDYq6^3Ybh3yf|r6-#8w}ojfFbw*)&zIrf1*ym97!6i<HqTU{%U>
zHw2m2t2AfO$U^kd;Qo%G_H+WgFL$+TR8(@4PVvIA`D5v|jLGzImuw=F8ZBPNphev|
zy|C7~RwJ!y&VI>czV&(GG`6z6l_Q?+EjZkYHPy_dj792*r~BkBp<NuW2!TDEFJNDp
z<ZASV;0f230ADs5gR8Z5Lktn?XB`MTM!Vj*oeH`yy`&INA~=I|An!$QMvw#uE++`L
zVF@>U_N~BNwtb>er*2T?>GZ-)XU|-EGTMf;j8!-|*3X`?!~sQglb$;@l>syJaFX{L
zi9omo(_vE^n}vV_J!94F>1f}eeR9_M<HPjNa_I3npj?(A;~oxd72Sl#A|iz^nfn)7
zf^PS%uDvt%$>^QiGM@7((dudhbRJ-2gz}z=g2UTbhm4@;%n5^@H9Fligv{Munk$dj
zIfj`f#zgk|oQxCEn7acx!Tzfsf_EuBt?IzFDu>$ruT7t_px5+=tX#+up3|&Ur`>Zu
zFrc{#>3QC)xlBp?^fO<>Wa}QA9@Ti9<5FOD0bR9pE*d93ui!rm0}U2zq9ub<+Y8-y
z8mji(KScNa+z~4mJ97FIEnE*a<*1nc9|`v!e1R^T8+btqoW<lV{0W{+*>!`RWk1Zz
z6GVv#MZUAluiiOn!pRf_$;l*aHwQYbk;g>x%sK)?-plWN{2VQP^Fpl3I#l6E6M6e(
zFjC30<K07J1Kq>)j`bn~g{R|DKT{vEC>uG^dtE#FxddSbyZQDgx{k}Y=gYLWRA{(y
zSV=To_pt`d3W=x)_$Ce7UFM$p3PDHk*-U<Rx6<_6ION4IG}svL0%z@J!jK~}rV`Mc
zsrDSwlW&~aHEvw*U$Ya2DCd2;35mWLP_Um~2|_#Jhmr14I=b^HJ)o}@`#gq&ef0=u
z=@Ki#FD$c_-C(rcGr&MCp#S$8tVr5f#o6s5gg$fcnO|$y1<uyfg{cYci6Z6Vm|~vz
z+JPvQ=1Hiq^bzCm5HmyILvg#c{m#b`Ew0Oy3+;o1LBk@Hq90pzBnnut+2Ef-=U=h}
zIzMBkd@P;ZtRi`^YabbZT4^RRMEycI_?w4(9xd_Hji4~hR=XgxL})7Exs(I6%+%^f
zDCJn90aM`3iv8vpyJ`0vJWyDL(Dx&C)yWHN+&#YtM%&L-+fr|Uye01yLBk;cB_4uk
ztsMTf9@x_Rld=U*<rQb`;QK2KK~Q^#36t$dl?$11Knb}l<~lHBcJJY~;=#IW*>cZV
ze*Ew>mFxEYvRRw9LbDF}$>)A3;&{b%fn8XiMoc-U3rPg`L6b<ggBrcnHy>$@5?qBO
zAam4$8W0cbMYTp-2937IvL#$NY!mU$492gRxo4r-3ypK&<kMY)(O%7M?q=@64-<5?
z*R+mMTs1$(z+4=4eEO+6<&r~;48DZyN*SY|sj4<)=74a_ANb;3G-fb5F=M%SRnGXT
zZkp;|_2xe6YD0&-=*$Q12VP>PQZm8L5$5H5=#<PfZ)l&P!LbEbo}mZENV5+#qWVlI
z&dyjQwb_78^V(9-E(5IwI;70{U&ocQv{(w!U~JicA0+Rg!iB#P;zOaNc|zs0ae(yp
ztFlx>u-2^i(rWDc8|qit;U8y~Q;06VC5regS(jC})L)=zM;~H{uk6}WD9AfZs4yJ9
zSghc+A^EWEdG-o>`!%7fd;jUrDoly{u7QiWP?Cfq2HqvB<$EM2u;~qI`P>xV9`@%}
zsk$Bp&&n7}>n}(9*GU8*h)ysG1{uYFm*9!0v*u)h_t0NQX?hI17*~BuDTvV%B*=7p
z%sMFVY1Ct2v`H3js?_N-YE7%JXnm(8dYLZLSDdFb#OPF>KvNM<N|iU!DnDrg#hd9l
z6o<L6q<U8fnq0KETpfV*$S+-FxGx<^WLRz;(xg|2TGVYEUzc~F7LF^sM#+;GAMxrt
z(OwVqKNGQMrkBv!Q7!}XKT*jE->W|Qv|Q^Lpt>AsL3j#L`*{r^5!@742soIFM_U5U
zsyR`(v&hcM;7Mf3=zrn1_<)l1kyllseEP?Xg)sqxMWm)3+Dr$=Ppc*4OJ;e{DevLs
zXU^pg`;RHShX#w`E5|Re<1b=X?Acq=0taO3XDlB5GGW{LHZz5{^&0&!|FJsk6~)p7
zbzk{6WS@p1D24KO3CovU<>(o(rI{YjCCV&IDP@u;WaRz)yxo);L+4FaF_|J!znJfo
zgO=xA^tG-Eo8bdkn(H18E7|k?6W6q2G+QW|MIShI%DZvSs2)+<Sc(|BMMv%lo@C1F
zsAKY*9nw=dpp@eL)TDOdbz9Dm+9RT@HpygM>@*dR<9i)KVmtm1Kno(lApX7$<naYI
z2Q?NY`Rw%@6#K)v(q*ud-Y_e!F%J8PT}GxGL!=HFOU4y6+$dotd-jJOEZY|8g0t_i
zPZJWWtKz8y;jaYYNs%xkWrkn2Dbao~T5L*KV`z_tI=Kzl;+H9jv(bxVV~#EP-h50(
z!T7-PHwLVJq$KMPMQ)klg^9)^lB@Je#ch4v9KyJ@mEEvF30QF`E3)ViZ=!ick+hM7
zVTIEBbUOLReht}{bbWjM2W3AJC4UGmwwpj&8dRzdY2}V=_72d|h?#Yrm{Cb8#}qfj
zdNLj1*?Y1#j--LDQ!~Kos&FbJFKl0YP&Qy^w>uCm7)^m0t~F`8uudOn;iTW@U6B6~
z{1&yO=7a=eP*})qYm>zR#j|Pv7{=nEM=hJQQ{^94VzVOaAG0p>(>zpW=izI>OdS%2
z29}56ay>W@B_k&zpLbYe!AvntPTiWxFu&%|PIOApx-48)Zg)0y4$t&1qA*HI_Ox!f
zqtVFnJofEXc}35;T+k2u7|r3qQ?g_wcKJ0dNvGfF8A&#(E8dGZo+42`Xl-1X156IR
zZeZZi!|u!GNU?0B)w`-(RB}{Xa9jl#NF-ixYM%~mcS=F>8`_&i{5j&f?Ds~Xt@lP|
zbeFXFJhosghfaqq6xHg)``0M~%bwKEiR-wP3z~wWJ$+o(Gu;c3Mu&{L*oloZZQOo2
zHIdOUVERq>WQ#LSDL%61wV>1b`6SyJh^i%>oiVPNPNDUqZH&@+v#NK4cN3H(r<a13
zjqeNsO5EY>gNHuQU`Kdpu-|KzBFd-}Fb2oVwun}T*ACgadY_}WQ91{#zMopd8=dyA
zeet}zRRpUpvxR#!90Ok78sgbi@ItH1U@bE}pe6pZxm5rAx)csKxE$l`XN{iAu(0Y1
zeYs#uY?1aD*Ve&PhT1r4c4%e1d8LSR68Vg*Fj5Ap&w|3$%pjoy;K2iC{WH;Vf+Sse
zS*}`|Zg5i1Gx~qtO?u!inl-L4(wlE)pw=(Si6Zm9^r@{H+zczueIHWSoQYHTE?fVu
zsw^9e{Z*Lo`Oh&;J%=OXrIp`GAJC?RFVZ!+>n@;Sb5Oi#jPDb*Ra)1Do^#=|Q~Gtu
zQXJ*a=WG1OHTjz-`8=&Lgi$$3zLdvFV%t?FQ20B&iK-^d4DFn`sAyG}MWXx~?0|WL
z?G{Wp^_D4B@XnFKw0^;1bHHGu@O^dSk2a{fB9K}IYXAQXF+LXYcc6k_`XHF(dyl5W
z4;9+v=*D&C`;I}fPru6DgFQqz2=5~|q=;NfQj(^_Q#B-4PZxe3#R+J-j;i>)g((On
zKC1EZ?jhBL=cgNA%Ir5djvi9{asop(&&{H~rLDXyq=h%8pI(f9C>&CJ;dVle{Na5v
zyE>l36OM+u?S*|hVozBdHcFZE$~%;C;Tw?%PXK{Sj-X&WPQU31yIq|efOv4x+wi@j
zQYVzB-IVj3YQ1m$a!L9kVUw0wkY%RLK^FH(0`3w7C|lG)M!J3$Ecrmo8e#fEGpmeZ
znf+}mXSWaW0+X8E;Urr)JP#c5`UmIoXhy7vO;8EdT#sxbD!3fTzd0;HXpi32#YEy#
zWk?>Fc{<hiJfHv%)}_jA7R?M2H^Y*-;BS{k-j}XYb}qjmFJu<GWboC$VbNIh<nw(9
z92dz7S<w9Qa_q*jn}z_TM)2_pOasHTD7)@S;tgrG+QXV_+(W+RLJ0?A(vaYrkg#(H
zh?KWVWTuf{N%{41@A5^zbE3xHU^`}LBu+k#($A~`b?2kprDZz$9~~)xXCPQEZX=uN
zHE0WYDMYtPhZkQi@lB*hz+nil;j)L37zmrM%F|z&z=s6Z$9qK=Uf|Iv#Hk@>F`^s*
ze0^*&j6^IJ-tkOie1#T!0p{DBioWB>jp;=Y&0RI3ayLkhY_tNVct|mJK+B{FC)%|U
zj)Zfb8Jbd`#<{XQyT-cjD$Dqc8@T^^)WRE!U2U4f{c&#!i5^dr)&gwPq4q8uY%z0N
z97ZG@Zl?_GL3-Z(TU+QxL|3c_leDhQtCmsQB4pXQEylz@HQfxBvUl2=(OT+qx2Yzi
zbLFGv0asl0;ogR$KF?^RI7UCogaw@Syx)IVhEbA1&@`xKY1-Mk;oy91gH@a#61_zQ
z)0xzKFR{i!-;X1@$r+d8eys>g$UZ*8KAmgI6M|q?FDf+m2@ZG?7I5FcoaDO@3ma==
zn!51Bat}*!7yY;}#jocaG4&JbHEqSCIZP~`)pkHDzwz#4hDQ4KiGr`G{ALj`3NbD`
z$ceNy`#nJrJJujkire(BezL;l*=2fPe*(>N=sv?1^3bvfQXAN3p~KX$yk-9Q9@n=f
z_<m-O86?Ouq7cwJ+5QpPfw5cVz^W|q{^eU9JsCXb$>l&SMmJIx@r&lN2(dSbdH`O<
znWpZ2iMql`3wwhYv#qq$UtFpN+8}T3iA+WeO>>b)H^f1vXuupDS1?Ufy5&lMF~1Rq
z3E-vvvX2i7sZ%LQ-!ko7QLFl?{2`s=B+gJAhJ!b=O@}2oy+a>1Z<Q6+*xBXvPB}XL
zGxF8B3FMuS$R2am3|;e4Ttwv9hs0?G)dj1DFykE8vFmxc(bll?T<iqyf>{5|Qye?2
zq{&}<cL8|bLvk;yPL@vNep4sxGt-k+pj!2sF-~Y0`1TYz%v$0>k+4_jJ<Vm4)3<@M
zO^)$rj%21&6oo|adFUxeC8%{toJqd;D>>Ds+xnRa)TIVvvY@W^%_Qn>GrP?UXtkML
zj0$JE^KD|JY1W0`t>&`|RmGZamS3A{57&*7hRsOs`I63qhAi%!6MpuPdB$2yEidso
zjxLKK@u(1|S4{|MN2F^vx97+RA71-S?I=+v>@vGFcOZpD&J}PrKNS`LPd%ZuKPuv}
z;dwfovwXE3wAPU&^D`lbB&JKTe8zF(iEDSuyvCw+`lojnYj_f^F?+vx6l28R9QuCV
zZmW6gOz^F2hZFCTSmWRrORF0;s&%m`geY3k1M~s9Z;F%0I#<oAb`*;_id_If1*Om6
zJTGBFgUE4opE<J_ZN3x6JZNF3D_GL4k>t2~^y@d^D4ES1O!@MLoxrVn(#)_{e4dOJ
zKYrb7EE!&*3Oto3UoEmsS<1PC@^eNn&#>XB4V0`~ss~~|&RZyM!3;Bff-;78`4AcH
z<&Q@X1d;KWDN!7Wgpt5}$Jt_BlBS!qfq1iY-|T=rIx-(`LYq|J$Y~XBGI){Fic6(!
zEj)55>8BM7?cjB0%oLK^HLx2tc}PdQ24;-lBOD30DYRNl%mfR4+VU$YEAC%nP5t~9
zxOdFJY;kX&kj*ZP?7KyldArQj!CKC3O3aTrNUjHFK-xuZCgQz4R$)OX#uA|z8;`~C
z2F?B@9g-6M-Anu+Q^AKMJ#yl#*>@Xa95r%j+KHOas4JJ$tgjZk&Wcld!eHwBY!a}L
zFo%-_6MrX1(?g~%Iy*Yu@UX@`ruzqFTXZpWC5g)asRzo+szn#Cef{hqM45HcIWu&m
zh|e1XWc*#ZkH7c<H2Csy96MPJ(SaQ3q3s=T5@WtfMT+nT;1x4XD0fbGQMVa*reRv6
zc?)W$a-*`0Pu?V_B&mNu7N`V7SN;ejjzT^G1zbm|J31psdQ4-OP*6UmV2Yfg-Lh$e
zxT7Qy(l8S0lx|vMPoa}Cq5bIg`;^APtby33ypJBFK_TG<T1_w}&smD5Jg8cR++$f5
zVfj(^t)H_0Q%|1aBy<63H3ZS{Ix=S@_nyCmqnxJGK=OqdZM8I*{3yZEAn|ow6uHb(
z?(Nq@8!*shhr)D+rpZ3j0bR-W5d<9}1gO|5V;HYBpmuKOGaSBs8UfJ3n#oeXF#bpp
z2N+a?@7&!=X8;D(+JFLy@{ynqqOv;<R8j4VxvnoKgT8Of+Z&_bMH2%V3~-!w+dNVP
zRU>LAra>3+sDWQX+H6LlR^ZVoi*r3+J^OyC_N{h@f_WJ21uxTo8%nai?f<SH`+<AL
zknFHE8%<Ub2Sm%LR}2+%QEKI+)qKZbD_Y*ddv})@x#B+6Bii_an}*5>s26L0X9Gxo
zxo8GNUYSM@%7oJ+>5ry#;zxj0ASPx_o8v<lODB{SmY{szIn}`h!XNm99=Tt_@Iv3q
zVC(JBYqI<)G5-(K01wGHJ#?wx?(sbm4+XT!^X5cr(cALw_M0TIbK)ci0bC<8I~|q&
z)UIhxT3n(?wi5oXTxhLs$W}=3SsDrWYmC!8E`GeC5qGy%Ne6X?$+kjJpH>qV5r@+M
z@`M+M%@eC@`S$3NvQ-e~Y8VM{`OFO9jZ665b4Syohpzr_G4$^yQxmKAyBOSTeiiRe
zs$OVg0HcN~XepSh%?Q;D804l5z4A)CX;pKC#|ev#fI>o3%B0?1fXR#7nOg#+A}EO>
z2>-V_j&s>asC_TI5)EPN-wV-jj#$#1AN`i{w|{VsPAl0Q<sc6u?9Ocpe(y<zde)aK
z>wQk(oqnJ(*8#@pWLITEEGGPaH^^LDd)5~;)>nz^a>al0;Af_oW`7{B>t%cO!rjH@
zj8ep}=$z{Xb(sCj@>GDeh*zd~GUgc`P3QL;;B|@+jqK{MKDJ?&vMK2~T!4k7=4l7}
zENM0ml(MZXktdcP!Pjf0bG9f2Z~%h9e^!dLRv9C6tOY6(7J4rB24_uVEyo)Shn&iD
zW^{JHyJZv`$PSS-Bp4MZy1wu^bvn(Pla>8=a$$>5Tq@d^5r{V9WDl@bt5eQA_>#{|
zbs3%BV$pjVe9S=P-KTlr5-(=d<jBIpN<WFq-0ym%afw#hCY$A)AZFIz;yaKhMT?_O
zI_1qY=N+HL7NX8N|C*Uz@Wk8PFnVvkGD=;l{PII4;}fWjfoqG?Nk+Ibk|?@DY7aZE
zYi*?XGUF}>tT9~tK<(h?J1w2(a#E`r0w9PQu<}LZF5-s2R6JQynMWGZP~*irqagAl
zO0AzZ!#!#RO4sehZr<~L{3w8%!UlcO26~~Y6wt**S>~4T^tVN$hC&QkkJfbl{b|Bu
z^D*9tn-Pt>ZneAYL(r8I_ga!Cg-Th%&}CDklh3eBz`C{MbClSXC3Ei{1H$d&ueXCX
zVAY9!=Q3kx?R3kkloNwDIXzhda(cWE{FSF_i+kd(S%1R0xu4quxnu|fa`T|q6Q~Cl
z+<Z^GMTSqzH8$8UasS>)Q@3n&O#8VC0-F3yrOC<y#9KyrKxQi@vM16x${}etQ8i@C
z^mp5gwIc5pmaU=&S}G0u9#Is0dIn`vg|o1%i8|KC=h{CmVWc4kke0`$L^EaF64LFY
zg25BzkO*<dmHlqBv$pD`%h7}+*jv8KoUjV^^Eu`Epi{LA$}MmoTM}Qo%{Z)k&*6SR
z&GAQK3CLAj<8FiYzSr|kQUR2w4ob~97FqUdZESi&Z=+{bmvz(iOP+RL6ukgG@K7Kb
zB?aHN27+48h$iMyD7p6Tt$j@~@)9U03b0pYTZD55JbV%>%k3>Ikw`A1)b9m5%)-EF
zmEsKd$j>meXW@y7Zdi}=8CjC8GGpf7zH&gzyPc3xW%A2zaf8>eyXV=zKMn68Aig9J
z!^%s(91Yl2^KDfPShl_y`t6w?=*XeDZ1OY1kFClZ1U)}76T&o%GkCU56#vTy&gr>-
zO|P(n_PJUE1eXv5Dm1Qnd4`?+H?Je=*idB3rHPwGAo&RaJ?2nbFc;=Db?`uOt&S(@
zB9d$~8p6H+L&GEh=$x9e^$j|#uDXf=J5^*zw_Pe5yFHf=M`!sy)37l)kwbAj;pi-D
zna$h2O7k_Iy5mEez_Nby!L}uXM-yzy0l!!?>!a3*2`!I~DNAIvI`dR7=0@wfRh9lr
z-bJKas7!=^2UXI^o27=5yeOPzi&ZC4&WSiQ^RNbD*mVX@RpMiFPWz{r|HPnwluERq
zH9M=#>Z-0D&<=v!d&HG7TAe@e*6&-=yL7@cE@n>742C}Qhgq|aVa2R#3Bp6bjl~R0
z)>D}lVAqDr>Lt;OzCGC?i+!i*Wg**y18moL=><zVRI^|v+RwyOo%&S7i8Pf}O_Pdn
z|BlPxz!Ov&f%tn1&f%x11^f8<G9?y7<Kx44@AzU7%p(-Qa;pIkFAETeMCUZp(g4YR
z#@>W9jloGNptdYE1B?yP{RZ*(FQW&BGOWHvCDk55-Sl844mU0mA~SMa)rn^Ga0br>
z@91c&A18J#6?_ERsw^}QZx7=jQt($SbzQ*S3~F~{at1xhp-#fXW&R|rHNi9D<K1&l
zo^Z_x%?~ekC@vz&I3+hP;{HBSvnrPsX$Fn!Jw`CAo*{-4iNsdp#dKeV`CduXb8Q5b
zI|V^E8k5=C0fa42ccFDwx%s6!3c9dR<g3%tx(jY1MG4rjB`qknS&iL&l7iG0o0s=9
zKHLF*iN)Ho%npG35CcghI0bd!FpV*jb^bBCn424>{g6i-xhFzGk#L8JJ8HPxlEcdc
z)4}0oF)a8}vA(aq0@k#55Q&=iJsNOLNo1O;EWqQV1L`J^K^fGwJ&?WOkLyX3+n&cd
zRy$w#cRLlojuzu>3t1uq<W?)}f@1lco<&^vt252Mdt+tG4vQNj?8}<yW5)YcrdV4Q
zA1XpWRgGpS2%=QOm>ZK}&s6&)wL!?#+f~B+x<&AgH}Wlvfp$v+|5RyEAb!=J*QD!$
z<jKKXGHAiNTflms$dJMQHm%h1d`xKG#Rm#<@%gUwFXq$GPmcK~a{I7Y12~RnVce&k
zA_Wa5|7?kE1WL35Mjqcve8!mM#Grix)z0_ocaHzk8lc@tWmPpD*2<#fn~ycpfr5!Y
z)7sw20AwVk=8ik#+rW9)iDSZZnMrqek@`n-rFeh>l`ebV7Qnyt&GYdt`+ZGvOaB*p
zbg|QPvOsNom15?={3qBb0oLYxY#i7LRaMDeV)r)I@VbBxXl{MbL{)3YVTvaufuDO0
zi(#9Bi%6)GU-A?F!I`u@?|A=5Sp1rNlrszfm6-APFZ4tad$y~b!`ZBlpL^>(cO|Pc
zk-N+E70XDbEMkDJ!<t^eefjRi#E+Is0$fu*+<Z&F%*mSAXKV;zJ2in+O$0aH+A9y!
zgcq+esEGo8Sbs10X@cIMe1Nyh;F&etf3RGKu{5X*i#)23QLC><YcC4>>PO;y8NFNE
zNJS=keolTua~O$jwUOD`&&hTGDw5xr(ZY7$12Ze`*vhj#p8|KtTg*i*`^8|GYzLrL
zBMmrQn>9gC{%j(Sss^S_3}watk;TQ*@FvyZLa6Q|Y$N`{(lsw)&oUTTwxm-22;0(4
zNF36m`fe>Kv;B|?7Y8@G)#|L$OlBz6K|3ySOYQxPD`p3(%%*B8JhMjq%{#K0=^nAK
z*Gc`=w|3vzat|vuNjZmEwNrl&DU|r|`)%L#zC&+bq9*&Z$AwJjk>JT27l0Ktpf$%N
z_~GoUFF-sjVw{;_9UmVz6o6?GUxvkuY=B(pQh@WtjOl@3I7QdL!-9F){&<`S0g)KK
zRsR8}VgNYZbG-4;vqHENe5?7mC;HsorNJp*z8Z(+F$>V(-~h}>EfHwHd`^CT3~{2l
z(M7h&ZB6QphEnMS-|kxAaRSY3&w;BeotjZl8WIA50|xL489^ytQ_QX0y!<|Kq+89*
z{piJ~3tL+N=lrEn3J4@;$Viqk)UrZ{IByJT-?!e(=QCzcB!;NJl(e_sY5)Q?5T6k>
zFIm}`5yugusHBdF?B^Nq@g7kd#gd?uFAM7x<VEqZA3n%nS67APeFaFB1}<bUWyu+R
zb-|Q)jrAzcB%b&aGDZ(&UPOyGWy(1~=meY-kPODa<XjFEB>6C;IbB^w&oy3)8ds#b
zEq%hbY0#=Ray+XSywI;C2MqqKj=Zief(pO$ufn^sEg;V!qa%1TpD7jn8YkyA=LxJZ
zan$(iT_>;bLcE4ZQb+Hf$FB-MGP2|XMECK36@n1K&K5-Qw^S><5WfG*;9KUOj+~++
z*J%d-A$8vO-hDu1!2q-T3&0T1#LpQvh;MG~9#a(N?Q_MZ|9J?(q<;=1UAO=`>tBUo
zgCKUHZPp|dNttO6tHw6IHu>@M>lu#K)gh><$?MKXL+^T=0o4NRlfRJ!!GC6yf3i0Y
zy*kYmV4E1ly;(K3Iz?!|PrpK1myI)inP%t@OF9DihmupSyI}7@yNPtq)B}VkEiFey
zI&oWm&}2`%FsbN!6f5(WdfWpLq3LqgBaCTzq;{vq&X=_AcMMM#o2oUty8)i|pSR*8
zuuB~hNsm)N0G#lOVhNJ%6XEF4X(Ii%4P|FU8>k$^X^rJ6uUe)}2);q8{oFI~Khwx=
zW#4CLDns^jZe_^-HfQx1vCiVuj4fJ%*cwJ!C3OL=eWWkHA^k<|0sTqGb|1H0P1=x5
zDUV=nt`;{b8IfYcOt5)hFk-7wk37IO<%QbXe=#h}KdcJ_*!JAKQjmrN%K&z?B_yvJ
z+vkyz1Hwl>^B+ETaWqA*{@KA3L`(Px*awye=92<$#=YNCl08Abx*XcxvYoYDNsMRL
z`fFG7z&47-2;q{6PQ=a_z>Nd`r@Ax(7BdPmX1U|S=k)>lnfN|_0Ph>AdEVOE_Ufui
zSsHnLsiUsFx$c@-y65fdiF%xouT05;GP_ASO6Us{p=rels&;jh0xqw4M~HyG;69gz
z`|RPp#o5GFXV9wb!?Gx^sHN-ucs4x1IAr<A0S%D0K+2jvH6PTglm*@dH{V^U6fBv^
z`Ckhq++Wpf)doJhy``ZB-Y}vel*3%%*nlID<1uA+y5N<|Ty)S7dl_{=JlUjW>oni8
z4DeAMO&+@}1#8OeqKrfkU%2fP;sUxr`C6CyqCTnl_Dn53JMdcQD%EBSPx5E4&o1Xx
zv`~JHW1!`-jhBJ{{g2TwYkPCw1$Ns^Tb~uwtGZ3?M8#N1RjC^u`_ZjIv2C#fX$|t*
zH5%Xy5J@|aZ)YF;qfoClgFdz_Eff9Fe8l+z;8C>$^EvgSp8iN#bMNQuz)RAc6|%x~
zr+hv>yNZjaMP;!E+o_Wvifos;AKdr#D<?@2%~{lC5RWefRfGbm<Fl%@)ez=ra*FK_
zXybW08Uu_5)~RPPuYeaZN`7hVWwXts4I<4ff^*4Op#Wn<*L~m#GCeE%O{Y-uiHjpI
zg(D~Myg_Xq3Am{u?37Sz7$C+EWr~q-g5(2lp4n5n5A{*C_y`}e@Kr8;w!!C*&NV2u
zSfmI~<CtwsU92Y-u&Z{WX$q9GuiSw_x4$>NoSnnvBiSl>MbND!idsS_U!LdEa7m2I
zD+6ph3?a!~sQh}dmy!22tdNk{nu>-B^~j1YdDPoGMNszs@P1_JhWS-Uz}0ibrCYAs
z5+{EgnY;7mFmiPN6O)+0Tkc(<yFU2-;X4$KHCEvfsVf5eQH(payFHulD!JyjSOVd{
z=aw{;Zto6f0}ozv44ZE%+#JOI8Q%H23&o$9=m?Kt%hBcaxVF`8ETKR677dk#N+6eC
zb~y+*04XLi_vB<dgHlwC!h@LqjoPR{_FnU=_3-D%Pi3A%5)!vVP{ytFtr)2Yz-d24
zT|*@Oo(LG%yh6X7F#KNjjoZ`KNutvTl5(!#a0@6NC$sq?=ra6F=cynGuyPju<4t%u
z@xzud{IA$frjj6vTSI|e6a$8-bt-d?;=Jbr7aC1IWHeZ484M}d0>-baa+s@GBQ9Uq
zjWZ^{(AsqeUg8;kR=O623QD2M(H283$w^fGEr7y83T&_#K`8i>ONBHg0|ni*-*5f;
zhIB2(Qt|7jYa}nGn+jNE740tmY{3k9Eb_H!7v&WgprB7LXd>V(62cL7jUc&1S%fs7
zQ2jqtePvix(blfg(v37qcXxMpcXvs5H%NDPZMr)Jq`Mm=H>ISccj-Cz`|f=ne!*t$
zIma4n<ok}vL-+NOl*(CMNoif6$F9F(khokFuy?hH|KosuI0SKpvx)!=i1gIGjlF-y
zRmZ4a6&8f<p8g#>8gG-x_fhh%8zgHE9ffd0WEc7Kc6B=#cH93hKk%Qu!i2GlA2CtV
zMYJeK5_AapPagQ85_%u?W8uzg+@lVeFE7DgmI%o|?CC!{8~M#KG>yJ^$3MHaX1ML!
zfA}aQ|8y&j)*^_Di&V`(!NYytotIeb@1B4k!T(0|+V^|qGxR$uz8W?vH*=_v5^yi^
z-@iHjhloNFN&MH6J9X;w^H?W0YT7CB{M?`^XMpL#rMnDNHgDGDGjPrSXVc~MfCmZ3
zf}G<pnvs6Jl>G~zm(7GRc4z!QR(R+i&-F)t=S#9jUPi$4FO=-mo_`cQVUe#(=6$p3
ziJRIwv{$Y_d{7hqqAP?GoQ$-W_cr4E01nOfn*VZW5>Sg9PYl0mKdT)a!_r_7IN<!J
zh4k;^4)d+}@soa^yyeq12LBQa0X7COY+9cZ`>&loDuUznKlu3Hcnj8o(k@uDpbew#
z_Vn<-`#}7!uSGp@W?r4D8?K4aEO@l;8j>I<76(>KK>aW1BOw7&Q5rj)P_Tq6Eb-&g
z-{4BKAhqNB-csa0F6qCT>|ZdSloj(jJ-~WArf?M0%PnC$`%T5tVx@8~K-_|75Y$V>
zRodV56?l?=`9A*wEOZ?<>>s>Qy1O0L04>HpBsl)FY7CzCF;KPqlID^6-KJUsP|H!X
z{v&5Cy?eBSJD5MEb*R|LHybAD%2bn>R_!8Z335U8U!agQxqAPfU?2-Q1({9I#Q{3!
z@JCA}OjjnbQRkNTGYMO|7;_mdwYzFnpdElfg7luCM*5A(?RfLx;#;LG?Jp=F%A3>h
zNk9xxV?j0=cby@zo}B`w2ZPw^`5PE!LnmpQ(AHhhV?hAt3To^7(EWZQg6!~=f?~pz
zs=`&Ro~Lm<3AYlrU(2+S;SgdNX?zAk*nq_3Rh_P<amirw^mT#Zs~&?jp%>_Q1D<ez
zK%}F@`~ZzKyD<cNnYI%ru07t)c(kt+??0HhPle_xx+I<~7mJ9P(Rj|@wlBdwUtoNq
z2$?|9!}39y5K-auYS=M&z<$tiQ-l%JE%-fAgq;g`S|EiDQ5dvWO9|ngf?@B=d<(=A
z5P9@+|B|az6DI~BPVFcJc;8XcA!wXvHZ|fb+xWV_Gi7YnFRIsBHg;NnJbx#lYK5c$
zw-87k2^^@N3qdE6<WrF%<AxUIo9`_c;O%aD`T7@886Kd?T2~4YAD(wytUmvw<Iib4
zG>6Q4FN|g7wYLXQ!bqf$)sotUFspyw1dEd5P@rH&hhX_D<GFv244dQE{_K(9w`oA=
z^gRj%I0esA-uxBz2awP+z6n5WHB2C=5S4fGt0f70#xs#5N~A{Q*pF&0;gzJ@)W*kk
z9dg;qBvo|tdmJ;+>??qGt>ra6K-0t~MApe58~@W82CpMLhk4lOuZ|5pvn^b6xAHeO
z0<eaFTv>ETJdt~5^?v~d;UTH554-IRi1S%zYHn$py2ioqjR~=iwBA4pWmG0KMo4ZE
z5j%rSQd{%$B6KSFK&aF;B`a4LexZxzh|C^?43?&he8;Tq=i2*&+hqV<!vLJw(^O+N
zn|l{&Tk%mQ0nwYzWX+EfXCP*_w((=^aK9z&yEt0|=LC_yxaE?QYT<6x*y@|t6hdI_
zi>j|EPs?2v>N~KNynsyBHJla9=Wf>CGlfx;V%0M85#FlW_wcN}H0uMOwx{BQJW8IC
zNI+Gky$byU;SoXK{p+WmCrt0f?`P%tMLb1&etpK-&aC1(vbx;VdK<3N(`#_ITQJS+
z)A|omoDphPn~>Pw=2H!c2tP6hNk|<WdG2h+c>pb%mQe}5{Ly<qx*|b`NGt`f<5~Ui
za#m>wmjvln&DpxHPg(&%c*ZSw9XadL!aEvC`Lr*=5bF1V$E1YK{9w>Nwj4B~)?T`2
zHaqC4engG{t!iI@uXB?28q185^9HvJ^Z0rve6{XJ|3B7%A{QFGEZv5gpNrc9UN(NC
z^KPeaom-oi_JD8hxtx40$3aMfY$>O1+@2$=fUxgU+uJU45k`36$j?T5O9b?E)}^LK
zSP7bE)Bmk7QW1ECbqu|V)^!*wAacRd-UsL9fW&I(_|>c{G1P(-+$i~d-Am9kQ;Cpu
zK9b7>usKHoYvf~}({rxCgM>ab=6okjhX=lq#8MJhoxTEsT*`$A-)V^ZG+S0qJ(xx*
zyz^(23u!DDrJro2<B&n8rg&8jTdxl3r-379-;B<i<>H&&L7)4VYkopBOR}i2`y`sq
zPVY#T$`>emi*5>2uGG$PX!4wl!B1mBH1288gv_SZVQWXhiRI(x#ceW5(Bm!G<Z+cu
zI~sh>pUCv~2K8i(X9q<j$40}nkAi>E(>bNZ)_f7(zR!;D^|+by0&Mt-7E=0S(5IeU
z0HN2{L<d})D{z9(h(<>B*Z;9}&?Sr)8D{#N6Ww2xdF18@J`X;Tll$~m^LAG5wHsE3
zpN|{OoGM2zUeK3AutG<xmZx{oLa7(HT&FZ(NzgJfn^c2EwChDY=hoM}b%3|?qq)xP
zpqc*ibSo<$+5us>P8}0?Ndh$ZiD0ZjJX;T+c0i)#u;{zyTEJ;qG2`?+RR~}w=!CIO
zlENaqi(&{$gK1E!8zc<1WLG%mW4By6#Z@)Z2@i0OLVm!pz?kUXT=9;*7$0tZjZJJk
zV8MCusI3$(qzF%H){a4N2gvk9?IiUK>0W0*7_n{=nC{s0EnKm9^yFWgaLhaiS&f|U
zOLRva3#-H9+w>@=zI$s&LBEEI72{piyehdT$1zk4qSh{q1-#m2ZD>m!U_}lQXg5=b
zHM_sqsn%R;VDXLvdSpV7{t%5VV!OnJ6^sUIVLz{p<5thFZBl+wj|9qsK(~jJ*z0N_
zGQRk*Qk`hTxmXE9W~}pE`$Lrj56eT4DshO9@L8~)*FW2Xu3ezxbzF74@{RR+vK@=^
zX*y&8mCw9b=|?U3u#ZK~<H?!}Eim;V{ZF1J)cf_tBFSkc$D<b#58c8VwJ&3ZN*Kbs
zlYJHZ-e=55Jz3&KBL#-<0P74GH*_5efU2sJ%m>5}OiMi%iHEi0Je|m4@GZ;g#2wEE
z!G9H{w`CUnT$T~_;0%37!&UH^p`EbCa$GMpxRer}T^N&GvF8WW&8xqptxB*zEOl@w
z!1$4}dXLad_xy9=IL(n-fR?@T@g(%1>HJ$8;qz|giFp*by5F&0xKO10X8cx?G}AHT
zH;7FK|D5+1@Fhvh>Tky2+H~|~v9Wm&A3MZKtcX0SIUd3#qG_+HkF9hG|EeQC$zQT6
zkQ$V4EY&1)(D;bK42!+B$`o(?tH4DZ_vT%TcQW(^IS3Wj2FloBDGvL>!~}}t+>cWs
z1#o{JH7{tv>uOGNl^4zo>By%4vPHEBrHrgTsa+s1z9$^8jr5o7N&;nEK(PE-*;MQ}
zOv_rLw74wFK?GP2uN!G*gdsw_-)GEFn&RGG#msq5qV2;A2RwUWiexGfSk!(7%9f!6
z7k6Bn3W>G_YYX#ooW)9-a;!7J;EIua<WAAuwz{Iu(mDR3NFnyF?|Y2$xu6oFUu($n
zI?3e*vzQgwy4DZ<-0rhF!V6fc<t$DoYP!|r^4l<fj;OGjtHvsPL9PlTM-M7(-&?V{
zwiw<92{E8w?>c%~+^!issgD6Z^WUgxlte~nP1>V%&Yo%!Yhm(+7Wmx{WG#}QwchCY
zkv_Apo?J@nzi~=aZ8u^&rnV)7uTM%x|E!`uwpY7)RTaz3Vp1^)K4n-M&vM8{6`V#(
zXSM-<%&4{#k7>f=c-}yvX%aj?<5}OYMUcafP3ed!ptW8b$F__uuy(GT*3U1bwVQPV
z)xcbtzc6Jj+jw8RJRKKl5LU~4(aTE}DZ>J!WQK}W11{?_K4sD-LzkSO?Nh^IP{n3U
zXFoq#I&}Bef#y!9*4}bg-&1ii6iaUso@KL$5>p;;T#5$WZ5=raFJzEg3qrsOuXnd?
zXwX|_?Rp~viVe|uQj6Vb9vlhZwb6Qv2mYN4<i>KAL~O)eTdfauN2YDmF4)Z*w9sS4
z=`<dN%=>8KbB!F5Ed|s{etvu573D^o?y{ACPn^n6%$HH*sh!;f!-1r%sDom_Sh=%(
z1VRB0thPqnsPDIYRE@<ZFQLqKSju>28FY|xUyFgTtL^HdJS6ru%FOaN$07-_vXW=q
zLSV4kuUgqIs93k+d%&clpo?&9B<NW4-FFb(q_bWbGmuOsd2}|D^2B=5)NnXyS%%LV
zp%dRi2H!%Komh&k^8v>#9eZB%eNnI<f%Pk65ng~=aXHp30<<cZs8jMw-!GczCZX}c
z($}O_0WXz&HiguOhPz$Cy$<t&A68vmiLz(S>UYv6&*H}~XnnyP!ZqDCA??R%U9zj9
zIO_A4_J_gzoqF0+*j|U}gmF9!gO!TTd1GwpPpr_wmrh+>V>>)*xvk@G?;T7+tlt#8
zQ>p8LiE~Fm_oU})ilTVUI><YSuwu$D9Tkj&hO1R_7dVo4R;(NR^unDgzq@$o>U0=Q
zx~|LJ&Rm{{=lrh4q~P!lh*)!dV0rW5fHsBRU*|qbqNB5+_z;4S^;<>VfhA(gmfJZ-
zz^k=vm(_1w&(m(>S@XxEwuyZZAgU6p)fxt!PguEEF0FrMWj@6*(Xde&DLIt&PgWBK
zP_#6D>s_BsGYhPP&%6NJ1Q#eP`(HUVuGi%DW{+)^nfMJ(Ec*p4l+B!(Tx|yP&zAO>
z@;zrz@1yBoXD7uK=hjLvXwQ2qrT+%?Nv}8N+o~J~*nIUB(e4B-D8pvYmu)?TYPPe9
zp1aH+`RQdzVVXx~ei&eog<kpewR9@O@~dyo1eC|3JstRGt8lWwr8pdotfmd)<Q{tS
zsmn7aHQ|6my{ry$e`Hxxyhs)UG=+P?!V;9z>!UX6`)tcde;5lYYr&#1gLggD6B^&p
z0yQ>z&oVYlj6JdkiNm+98Z`zCJ{IROi}Tq;dh2N4W(Ej^-!OjB$1TpgcVo}LutD-*
z>d;w5s0f3yfAZPi{)w!Jz&2*nhC9RmK~hfpg**LS?u2dIm*gCh|9uGFI)Xwdkydk1
z>2d^$;sVC}o;_tYF>1Mtg*GKmD*AIW6H#5UAL^uC$@SI)D^=E<wRF2>bboh%{M*Dd
z>8@{u^g#5BY7XHAb|lhE^@_U9D(X#s{d%yEA=zm>wHWeiIN{A5<8ccnlG$>YDA{1p
zl(RCQD_mw)Bcu5|TSHZ{Y?=*x2iGkL4y6?h=K#<;AzuNfcUtgjk+=*MkEa!bLG<f`
zhoC3^$>z1wpps}gw+n@nh}=hUc@v+e#=8o~@_VV<*+m2=zk6UrPk1vox+(cV1=wA<
zZTv{oL>EJo%NUDtJnnRqKD|{Z*ZoaIEMWpj9Y?9;I~r=Y=Qj5)&P(cgCygKaqYVwZ
zYBtsmP(aZFA%B`iPAt9Y)FF$gOczy`Ujxb&1ck!3CiuAS0;RH653TV}xFKylj$&OX
zFN2SfHs1P_(NJQ{q-RSM{TDj_IA%?vbY8fbSVL$*GsFh(-M^*sc-2Xb+>)Fk{<!SC
zRXSv1U|sH@G=`g))2(D>?T^if->RK{t=LeWk-PhyoA3e6v;Ph<6VK4LfNZOLpdXHs
zLh<stIXiyqeQtljQGHjTG+i`IIh4O*@||&8@%v3U%b<+PJg@ri8pY#(7Jbh-CgwJb
z1eQfPwc<9b8MKx!>_U}EJe4)~KZ^2ceLrK<zA2=YQaTJ6JUH*cMkXZFLeX=C4}hCD
z`mR1qTMhg6z7BKt(@K;5OiR7t>val)V^@G=ZS7<;?tdl-{ik48Oy}(n2CnC$mT8KM
zS5WJRkhZ37FZjS<zog+=Sv)Q9N%IKNQ(RA}j6um=k_AnZv#F9EM$<^(F7R%Ye)$_G
z^Ci_Uq%bNXozx;Rnrmt}mARQL#cfM|-=@Rg`cK`Z;fo%x*+vy5M$sqlRR?VoDr+`B
zH*na3A2JjLt+>4y5s?U4G9?{ZhkNK2QdvlCq7iX#=<9ureV9i2Y(8|s+GJAe^99IH
zAx|{$R!cgv=X+&%y?dm*WY+UEw5bl(==i*m!G`Z0$i;gzK=E5M|GDyU5phRh%T($(
z8*z=e0o$rpF14Kgb<h{=|0N3EX+N+l9p7~jw6az1WKBqPY+ehA8cJFU1U-wnb!64D
ziN|d|JW!6^-8B1j2}SNc@l9*QYN@piMBM|Jk79MDGRaLJvCrz;HB)QU(*mr&W_}14
z)~_%5K35<gfFcLSCjC?lQxX+Ev?6awoXU9)c4BwImdbuKj_)zh?!is_y&Uo<p@jgm
z$y)9Ax9F@IrlRH>GRvT+3iywj*~}<BP<>>l{IqH1pcny5xxu|`azEv*T>~uysInOA
zRC3(u8HhgPul}to6uy-84ED&p4Vo}wbCW~p2Bs9t?OorO2DM2RI?Bmk<y7N~mGf_T
zTqbC31^BQlSu(QkeQK3)=p!nTBTaEJ+-^Bm)ii2XFd3lnA@8dVvsW$-=}}#WzkUv?
z&AN@M8%SBL%_4g}Uj1NL#b5xfFWohhG0p#>RxbUQe|%F^lhCA)T~(t6KA-d;1woO6
zN4AkiZ5rScG!gzrmH&Sa3SNNcSK#r~Nm@7kA(|!6{JNb!iw;Mk`KX}-ziTq}DScS4
zGt(vZnBFT(UFgOlP7O@FF}Hj9=^$))xB1ae$=%_2`bw1taxDcxIL6SlpORm>Wh-tN
zMKOLO-K<d>EfCyvU-aJ@xut+VpF~7^y=S>I$3Ixsu%BZ1Lih8@hWn<T`B;@^4@qla
z_5P2(Zh(-()uOt6TYt>2pfh=EAiAIJu;1A!&>)}Lvph+zwV+&>>Fr}7Jz`)ln=;*;
z9>6bv?Wob$Y*_QY<!~AL5hO2TE`Vusb{~8fto-$q7_Uj7F?g8AwYb%xZFAdr^6lgN
zSL>_On3cd^rh0ncgfrB4+c7jS<-M$47&Og%xiqq!6<0+(mt)b)-P?J6PV%Z4rXi?p
z`@k-(Pa)v>mK<ns+oIAW62TEn81)NUJ@KYJlIf(&>#`1bGUTV0dPDEb8icxu{dVAf
z@Tk`M*~L8N3=BfaM9z+-W6HfGYp=<U;Xtib6Lv7H(vls9?Trt=e&e(%uxvS~02X|U
zc%r1=*g3O?nEk3&N_|c3sk4CTzkO_VM2EUDj#y6p^0lMR`EpFame39+n&<AFCyM$Z
zj6?ZQ)<6{VjH6HO947rS&wJdY?Z9JE{bh)c5F8y);;*#^n%N*5j2p!6KUa&H5ym@}
ztSAEpLRHllR&`w=O+s1wSZ|`Rm(#6RJhil3bWHp=r?@75%5`mBD_h7n#srId+0S-&
z;Qdsipw(p)tBbZJ-MG#1uN(FFEFIgmopjAW8uWMW8BhfuGYN9#umP-=k0OIkPD|Iw
zY*7IghW0L!(W#?MW--emTozYl^Wjm2<(~Pr(sh`Qeh~|MAaQ+8J1jQA`m~h2CYxo@
zZHtfElMVj{Tf252u^r4u{1!f(PL?L@QKeR8Yx|Xz^H!oR48Ew=i?8Eab@M4f5x?40
zq8tYiSW{o(bkdMW@jxCzEX~fZP%?&16XV-2bM86nnPBxNnlGwMMbvJ!<67mm{N_}T
zDl5KvnteMN4e#bN6fnJ%q)MmF=UK2pSdR&njy&+0>(eeWlAM-7bd*<+H;@ZV*tNmj
zRNamJPL$OQ0ce3F{x8^68YxSsmzTxQwVhlDC7`XjE4v%mrP8zAKXh1+p`$7j<X;>S
zAyN9r68Ln@Y?;0-pqb_~fDX;1{9~GgUhm8NC1W~GLa4ndxR6F53A#dfp4%Hwh5D-W
zx4p8Ogg%4tc$x_;&m$TgDA|6!GL;3Nm1jQkDhO4TnIZkg0B<ILw&hPhPiJ&V@5>*c
zSl+9!>BA!!1ZZJ-Z&2p1m;Ra&WoGPhtKJDtEN_0W)a!9~9lZ3Cj+ULEjrW<pn%&Sk
zt5;S_m~@cIraTXr1n+CrxBUs?Xj1w^*@vs3YH|uyIL=R9ZhG|f)ic1IUrkyL8+$SF
z|EHqYs6~<i5YgqCW;!Z2Y;Ngkbl4Nk$6pcYP*YhnO)D_ioaP@Wv=o>rBv<l_u17x=
zwYq1{Ooc`^T+{mLwjbV#%AE;?k5?@|CNpj?j$aqd60pWp(rlEpy5rg02UZ2?)O3Eu
zA0?RjNM^cbVRcB16HwXeDul5ygIGrFC6IoOvoS*`Cz$2T0P(6DziGq$P%WWZa}`}U
zzYTk4&V2({;^dgMLIY0{n>*PuG9SxCU<bPB!E?!8UoO6;kz31O-3`lsJUeYka2z)%
z;%IZuRrD<cvH-L`XdcFKu=B(1!mFmfKLVL^>vKpr*^#~7)9r(JE5%J_gq=^_k~u>H
z5}C~5>G;Siczh~4F8Q-&L}m{{3AHYzbF;`&U}4Hs$L!VRwJu3wk9N_xMso!9y29wQ
zDSN45V3WxkCi<V~(dCL0zHizj%<7QplZMjwdd1YwA_sOM$iuA?G}lfW+kG|&f)Ss%
zOp8Gq(7*+PI}lMdNSn>CEcu=JUa&fk>6;2jCw>?=t~b~f{+ufxMjeCNd*Gx_7E@R~
zoaw$hq(@DE>z#NuSzOXxO3j9mOMTro*oIiTI`we7)_<f!buFbf*#^U~25lYa6egTr
zoYbKH(k0;%(4(~sQbY22-tu+SR7P4}Zd2v#&8+_g(1#CS^8WUO{~;_U%$kK-o-~kZ
zGDlgM!}mC3hcI~DIs_}&+0QmeU9TZxWs9!l59H(6(iQ6;w2SgWA}d{!3-1tjXeL4^
z{0zs3v&XtglMcU8SO&?NXSXafm`)jtK#22|EvS9ja&C>Vzd-i&RTg9--y!41<wn5#
z)vlmAEWXMQfpCk6C}U1%CIEJMVQ|^BRh#GC)F;U;LY=2mp7~~K`iSiiDWtrX&w_0=
z!+x}illd^6+igfWHRp*L+#E$sYcCFrLcn>CP1UtA<*Plaj$;|rEKYmiH^<`CUG$LC
zYKBDjy#}^U(={KfqYKG;64t-sr(}x#cP#$IkrK{|ruZt`UUc3KBx@JQ?0y#IL`9nz
z%_u_qk%KlnqaqfZ$P0S#I$>FbWxodX#y64$%s=oquTY7KNLsiEL;SIbO6*Lyrq{f3
zs=rA=N21QkIkjIp?q*{r<(^E$@`$EMf}r<%55-j*(^`b&1mCvy%;db%U*SU~Q!Xeo
zqVzLhdu4UI@s0>izWhEY;RG5BP1u9Y#>uj?SiWT^@S3+dO+>vP?v-`BCAM*@sCR^N
zebzrdpr2zVp8^&|6K6;Lb9?1kdu>0XHcCSabpAI}^E2fywnoSO1R}I@Fm$MLd1uiV
zv28-1Id?YH)!3I@HCKJxl{m4&HuR9+t_c%qth-hVZuNX80L=Y=OZ)TM1wa{P<!qy+
z+5PhbQ-V4F4b@F$_IxalERL9~dt#+`+u^xt+NdRB5<y5eDk6IKVx|nomk%S$tjqP!
zcCc^8=khGKWLnxYrU{3vJ$I!B-I*r*!}i$yWy6+1uFFKKQv%Zvr|jypL(LK{HM#{$
zIkz0$BiR@`%lLSOV`@nLQC^vxqcQDfeR5w@J0S^T8rb!YQ~D34#_BVEcX8kGo9c#=
zzc6p)*%s>6(S_tT5}9g{c|$Kmaff-45+IgvvJI3K(qg;MYIP}Eo*vO9klU0M3Z|Z3
z>>`x~FKzZJ`mR~w1A5>6Q@4OiSs}hlQ7)#glg{{I@Q=wvlut#ZWM$;{$vqU-fFrKo
zilX<QqdE-aVh)njCdPb>%^d~!zxf^+A5D*(;yQbktuDYK2YK1n_MPq`9btITc9ITX
z5}96a<Ur%qw~Zg*F0N4)C>~vyp~Y(hR&B$lMj&7<uvmT#cUqg|gE~TTqO3~4VURsj
zemC*aw<$5rMg<f5;LI=BQu=K^{Y%he>j3Sh2Im5EGN{9xjjJJm1>=kx7UKs&T&I70
z#s$<tn-FB2`zqUT@4K)x43EqiJ~DO<Fsnf5W1wK+AaM&pS8~TEjdqj%wfC7h&?+le
zSY4eRwQ_*7fYKk>KD|!Ye)#jv`L{(y7{7_0_982bc-iX4ag&hZw_kfU+z2HOtCa45
z%Mf4kB}9NaItH-xjNkp01!ni}^=`+cdt%o$&^ro$PKMPwr+=T4L_1LVhZibo61q`o
z;{@~GHv4~`*iFfl<<ySf9@(Uk_L|A*8&A6L#WHxg*SF;7T&SSq<SIdJChX0WZfDxb
zdG5d>K<gf&=&@yE%Y2&WnlVkm5^)sf&_*yz!5SzdKN9zpOSNa#dalTNowN-4OvGse
z>Hx$QNwYEDoaSJWMi}A1QtmhU_2Xl+z2J?xn1mNI;9G`usI40FyoPdm7L%TOFGlTm
zPb$_SC6r-)TpG1t{^410#pZeBwaa>1Kq5lY^^CZfo!Ue-x-k-S`2`bF`aP#7v_Xja
z2603KM2h`FglK#&&C;F{JV#k@-t3ox4TlXsMp!CMNG{P8hVp527=R;pQ<bl5G5&@b
ztop3FIbnnj0e$8PBv(fhWc=rdTm0jBRHow_N0E|+OBvU3^AB!u=#pKH$9hJ|px2V2
zZ4RNByN>(}QlbT{nqYzx;W>}(gbvit2@>Kni^l;B3)m>0&#8XLl@*qss9M;ZO&c;1
z+T~O?WwX`^RaB84F0#sQ7)6wpynXwZAhWZw>MBMlja{R9QY(>ASsN~;Os{n<eiFS5
zDL7$3MtpO3h8WpTLRuqcR;4$fcLLClosW^!msl>pN!4y$&~>Zi49jSB#^hKCEZ%Zj
zZxJ#JgR@bvZul0|bu<1<&?BwI>?zgjyA6~rYSqx{&dA|=$N0`CFdKfF*o^~~dE?)t
zvQl3$eUGTQyK{H`?5l1;zr=@!aKI_(k!O^I@Qe=y%wKkzpstxLVP0i;m^e(ZhW?9p
zkdr`IEV*KtANjzLT)ChJ40C$>cyn1N%E>IS*MFC`nqIN3R;)Te_ECqPJ3j~^m@C=k
zXZ~)0ToPZgt~t2Ni2s%82d$a7+|jK3GDN1?aJ{uTeL{y_)!e$`62UHAg0w26yd14<
zP`gHQ!&K6*hKt=;rq8!V9rgY0DUYfZ(guU%&%?TvrXI{ch<bAKeV-tb%a*OuGZ2>7
zk#`262Ihx)O`vgHcfDN>8H(lYN;)UKXc^Zl4>~BkD(8wD%oxqD7X+B_HVQFhDST9^
z{2Im<sXzJ^or<ZB$6Z&nw~7>RO&Z1x>`u;<XTs*i>GhU!q1S5lz87Yon1#5LM?@gg
zJRCdB#F17dYiL3A91gL*G{|ffs~rv2$gB-*lNiX-!WqCEG=&?#J@hAro))ehwRqAz
zl$LMFiAh-3IT$B^LvDZeC4?~pAwXDu);ttZvT@k8F)y>Bkk|h*=j)p!B)GmG7!^cm
z{FF%hn)`Fh`NpygoS)Pw#4r;KpqFJUb!wwCJIw2uJT<QeYK3OUAIe4(Pra^&%<zn_
zwS$t6eGpFm(14S$rHuT!kXnS`{_8-{bqebGAu_EN$P!B9lhU)!;<TbUb;et>yOOEY
zTgFPSDV4%C;B?0O8Zm1)J;;^TVUl&HhVJN(z<#J~xEqq2=YjGruTGwFQH}Bn<m6+V
zn&<V41WcJeg;DnW9INX@_VSsQv7YluJ(HW+HV!=YNL|E$r%RH(kHH&dO~0yFyC@!7
zJL)WE3UG1ZDy`(~V3!MS*xK}7R%smGbd6R?VU3q<%i#SX=Z?Dr&#T@_IKu8h*~mMr
zmdQKKYHgoL(G`j-VqjCbYF)VrDH&Cn)$h$~eFOBL19BP+0sPOzFB2oDgeBD$wi}X+
zeBHR|kSbUC?GLHqGn}OV-u_^!wi)`QS#(Sr2vD5O2b-X32A|Ys`-;cCPyoxx2;}dD
z=6~I+YM;%rhbkMpwOP9Ik-+%pB0g_pjL{m?tO|ppfiDAzoHZs#%v=O@iH|Gu0drdK
zYtrUxz8-V?HnHDh$zH-A!OR7n)szz#vbp&-8V0?7xjBA9k^2dIp>ot9FXDydTNMzV
z7nr-hP~xjIoAaSEeY!3T;i~7Lj*Nmmfcq)XTv#~ZTVu5-(6;ZsJ9*xOf7i_NkTL!h
zfKMtEK!YU_LbEFHY=XFMnkWS6{@=|*R}b&s`k?@cN#5d+1Qv0wCzMVBwqJZEM{#fK
z!Qt|dM@Cyk5b~DUhEFd2`Fq(P(oOmdlVZ?R70`_xgwUTyUigT6o|a8Rjv@g5*Ir;4
zW_6LPM)HUarj(6}>{ps`W27epGH?8);$(Ll7@^qlJQ(pdA^4{03#7)M;FSpkaik8v
z7XMN=^%D2AXXtWvE8}dPltpFqxeToB4BQG+c4|qjfr4qH9ZWsSE1*K(+{NcEJGM48
zj3(Mrcrv+Ja#2U&n^{AJ;WZTO_OESc6|^g8B;m1?)txwrQtm}<SToMwz{5Dnm?`bT
zEl3+G_jk;GA+*8Z2Kzo9aOZ;po@e=nRT~^O6$Cf`KmfVXWSDy!MxGh;n7^Af`Cl|p
zxn{GFX}YfC+@fw4X~u@II6mDHEZx3dlf2En%AlNDAQW~atDx<*!eq2I@8ZiB9am$;
zuo5inGx_oDesiAk{y$bh{o?|CeGFe3{P*^UJFPS7v;xWf^>p`Uhj~u4>}lwh%~K?P
zvbPXu^DhCAF{1{7O>e*c5_p}D3rQ#(j@HY<eE?{UU(i0?@WRGvwhu~@<B(2ahnLfO
zBj*zU&Y0NvkX>bdA)Nntn$0ZO-jmb;@?`18u$M*bOMeI+Iu~8tcY99;n^eP@9Nk~8
z2L;#9V1j|(3QiFB!*E6mK#cTb^OY_j;&~klsYc(33CSC8id|F8dwuCmj=4hC^#txG
zDCE&fg-aKX>a3lZLqL$HrWY@yC;(S#OheR6A?O-<;=NsMI{A{_eEX1x<tX@OT(En{
zd$Ir*{1re~9f0eb=yd9ZJ}iF6h252oRqV;AlPLze$-ynk5f*`pJs-8Ar(^+m$Q2Hq
zkm7PFr7I?u^qD$xO;~He2vzyI37U3=K#aV_uZ+-|icj9BFFc<55w!J^lwochA4Aej
zR<@j8z5&DDk$m@176C4$H}`!^L5!Q$35rLTo8KO;mP>rS*44tq<6_L3lB>m0N2(~!
z8Hq#Bb~X40s%!3-_&}xPV0yoaU;CYR6Ud`|&}R^P+u1eGkkh~yYMi`AxbM|n)`{!^
z;nO$jx;l=EPiuO<_qAw$?sK-)jXj{LH+7we+~oljK=U+<4-=(JH^&_?MjF5Y1feto
zgS*_{p$5>IoaRy<!P<<R#&iFGiVhP=*B7RRqE*1O5gBs~TE@v?K<BKT5cJv^6A%7#
zD(ZGMCjKUyn><u9!Wb+%^6sw;e+rK)4BY4x5b@nu=XS>cQmbeFi#7OMCb2wFZZIi%
z`s`pKC_|)HA0D_wdD!@of07<y;4@YjeC`YOiILkrpd`kp0Y4Irr)J2Zdy8}_4;i0g
zGV@!X(*jX2k<`mMe&?g4SrB%<{)PqCWvA*4;mK6@`cBMsr@SkZn;g)CT+4VeZf>c4
zi@<D~7p*5iYeD}}BnTM1ZL<;?+}m{y+&!>=+e2J(d2s7A@}|g&3<c81c)3?LaJ;xr
za>0k6fxA_=UW{wBLM6y4D61;yd(*N9k98uHgBPHqZKh)##?F=|HV5FdQ+llRv4BXd
z0GNn%Zcu2JLUAHMGsWYX7IIuAq&~qzGWr31*EJ9t-gf~Fo_$VDv86k=@pLLQ))J3p
zpjq}SSVZ~T`b)Brvjmffv72@3Jjb6UWcVjZB&d$Rq=Bk$dC@n(Ha{xRy2O9$3<C5K
zU?2cz_RHcvzbNj(aw2A_`fU7xk}5Q;IK;g_iS4Jn43%2NVF0P#nvkrW>|1d2mWKg&
ziv)O|8dx4yp9U9#)(yOoHNX*|<mTnwo%GyU4a6GQDs*F{_p<K5tf9k?c9{9z%=Lo`
zr*yJls=bsoRdTIr{PUv%1(-M1K@;*XrPGx%>&UY132qq==TV5!Fy@LqMF_rwN-ZF9
z>~j8-NEVrmn&kfW9B(%kPl^vIQ-^RmQry+fGPLa0P%0U2yHrB;lX$gCw(J9iAt59Q
zYsT!n_6qT|R-at0JbppAu&W7INdQ%XMJUa>u0APM)w;`+?3T?4uT7$dKxK6jkYO@0
zM-JLxov6?VP?H{A61Gu!PNx=Ka#TWm@PZkO46EPAbx?L26Zwcnj|v{A{n;c`pLe+|
zW2$p4!68Cic44RF3aejjTRl?nK=4j0C4?6OTEFhw?6;o$?LGHLoHOB-v-A!vNs*wO
zGz9-}=&-E7#yO6+q&wzDG@v>673W{ioYy{rymFT}M!6!v=7e+5NrI}f*vxu)09G-;
zc_HCp*e|C+QNg;y4C(^_V_2jj*E}*9CYB`xoEw0@1)l9M0J|HE2!RckRuXJjD^@51
zDlm*j0`U3%J$&VD7N*L-N%v$SVo`xNDrwXzGXuL=DLqBNBdYQ2Jg`d>wT}gF`X4;^
z|6fXCJyJxUS^tDF=8TFey&3BDi7{e>oYh(eHj5J{{|+CR-!}@ks)wArv}kf+=iz?m
zBmNl=OANGZ8lK_-YbPK4dvxB?x1x57+T9I)pClj|bXW1|F!#L=0u<6|$RldH?7~o<
zLBLqMEQ2(2$u#blW@wnyt;Eip_*m(tyUd$ZeaoTCync-Y>R)c^3-3GV^g;C4KF?4!
zg=wj+2G`G<5V1-<-Lxkwp(LY!!F|DqCdKykWbCuym1M>GB6nnX^gS;>F2z8<pq8*K
zo1-8XMkqkEd0=_=^->CP+$HC5{KIQ=bn`48ldl|!Bnr|u)YgS_)Hp+wEsGAZbK(ue
zS;X#y?-(ChWF~M6%Skoq#`q$E2q<oGx)H#hbZ0hy(pTR6ae7hytSnket#8Z!d*(ez
z0+m=h?S}`r5j&EwCYVc#BY{f!IEs%yD`{yNawUe|%~$Cglz9>+%|5gGf^cFLYrPEt
zxw-}Z55tCwE(JULpRk9o4QdC;>qY9<+f@3~l1#ATGu(;g8~N{8D0(fVyfF|9nfKJy
zx29*Om@4R+%~K4mr*J=wJ6TVJk`%uO#K5$=HC+;y1j+uZLBHdbBPwrM5m)(6vNCuH
z#+3D@-J*~D(#_>avaD0oTUOC%ebr1M({ftdf~h4FPf_c0K6}MD^aoQ6_1iFRuQ=fy
zl1O}?q{-8muh%9$k^mxO5#owJ3&(95E>%|N=K0KsupdaAP2!fyHov{8&3P$nq<3}2
zUy357mboHn0p&#c$auPGh$sU3V9x?y><SJhsw+NwprZrw_vou4WfCw@X^sdAY9Q(v
zZhg>6Y}qDWG4!0lym<+1t)JQ8HKJms=q{W2UB9yiR(ux(&64Y*vrJsY@I2`2JrdXp
zl>~_)0-Dn$fx<NeA{@AoX$ktydP~6J2##&9Se-1+uUCHRX&gfT)7d~RcZ~Y2jN7=C
z`wdAnAZvQD$c+|)^*e>^EyB^Fm$zwCrA7YTYOc3eA&IeFQDeGRQT~c;6|_@s%XeR6
z0rf^@%TshVd7IdI2&X`_My?CtfCyOOcYV;&Smy4XUEZhLw8j3qAs~lxfmc7#dHl}v
zVHGFe{4Ui5$o+XlYDawZo!Jz>!YSP%`j_sVZ7Qi`sbu3vhO@X67Z13sMJ)URG4&qx
z{A=Wus8y}y<BPp)XTdNH(Q920NZxOH`Mr88$OO`kC3Bb-R!!TF=qRH?8v%;@3DIJ3
z{l#2G+9W?BzrOmox^<UTbl$F3hDLfns~-Xv-Czluj+JbNqok(_v@v&V^FziJml!C{
z2g?Hwg^pd0i7=SA<du0O<JkVWws??u`G`@O<02w6=8nAQJV!WdUmbjn#H8!MS~2uw
zB7P(6Mt+VdSY(JUjqY~uBzm3wMXl!f&$eUD4QHm}cKMVc_zL$3rt4j~P|;*k^ba8u
z`(EZZEWf92f6JIrJ$nH9(fukmi-80G-zxs_s2?>)u%7h4hc{x9lUx4*8|kXlp=znk
z#%gLYq0%>o`_w)cq-JZ!)|}sxmt`E~J+0NDZ+mHke(v}5hAFz>B&r*#VOai)>LpA^
zwW*MGWYu$W7OS}B?&${!0B)J}9$6(QTUyeY=9ev<4w;T+_rXPg>T&xnoZ8E&Y<h)K
zJumBmIAuH%PWlvt7^_N9Qpqh>PBq_Ng1SeCia12J`iw|;Y=j3eBMV_=ClPq0-FNLM
zyCiNLJfU#O{Sh$f$E`d(?q~TmC5&jkGS5&~amsqn*jx*+i_ZX!wuknOkaBg0loq1p
z(mqFS@2b<yWWO{UVV(R^8zJA`9j)~rSeL;9UL0zt$vY{W7pjX_Y&rR*QwEE+j=R~&
zJJquJl#bhs7l;_v&WO}a1A9Vc)fbTr_O+k(uo7zvzHgK==p$i|e!OdLw4j|7%YTad
z^+H!Gy^dl0)z3h`TV)=j&}!8@(hJ#6F~5fB37Ewlaw_-((%PdPWJieg2$k~Csnbqj
zzx-k#S=?!crA6WpPxn2;sIT^25>R(uIq>IWLS?+*N64iDyWh6!d@46IG(+x^)bB1@
z0eXX?!*O?E0(biMm*20KN%=21gdxt$x{ceUQY#srHobR={g!nY1W2Tt33{cKn`*uf
zr;q9)vpD`qE(eXToX+no=x0E}&U(H<+aE<RdFQvpH3&_j4mM%KFKK6QD{OJ`IxSQ_
zAx3Nr8}`CEm;OawA^Ib*ak|;I^lD4fmItbpN!&c9$Cl!&rE#>hZYuTS>O}*^Rn;St
zmO_Q%8m2=(3GzPE@wXJ2Y|C-23aoN!eG{t1Xw}NahyINW=X8~*rr1Z?#WIOnMSTy%
zg%pAOaV@|cn)3T1*nwj2p0c)<O+B<sER=Bs*v$RR;(V6W+BM{vB~|Qgn9^#?=eIcz
zec_8UOf7I_m#ZdgCJ`_A!^i6?hB)Q5iN94(9vjZBTxhO!0u7#w;(S$GT9WTdJuxz#
zqAOT6&iAb3M;b2YbfUx9sV$Hk<>5bTNJhE#+k}*hZaF-BaJJl_*+`-b6}IXaUaD8r
zmJct^Sut@<@?ooHO{<XPLd<#Gya5cdi`1EW%dKtWRQV-<yee8z!OGxh?0?CK-}yj#
z&xPl6jsyS*fZDF~)90=&`&qNCKb#F@z2DcM<U!Yrz&{bAvSu>nc#=%}f(_KiWL5zf
z0|e{4mYyS--qVAgagaR!w8ceO$1c2)!`5~jkz@$Z11S@F0*q+L40EoJOo`ORy{nK|
z_Tzi3UsiQ3;kNT;esGVof`1ih@;eC^aq#YUx(ElTzW1gBWz5$f{KhwazH>f^^`Zan
zfI+=r%(Z%3rD#ayxH^W5RS%$WtTI&MYmkv0jqoZqw%~aYoYa58*9&d>3eCr103esq
zGY{cGqk~<f#x&7a5Mm;qIGN(*0?x;M!?Nkg1UCZB)Bt1Tt}a7ECf7)%iX2b}?bpY<
zCIAKc?G>FQI27Ra!42-V3Aw1WX`E&sH2cC!ZFnuCzeScY3h&`6arvY^*|;32stumu
zDJKmsIa296#(8UOxa*WMXwY8E;cRp2{}w8()5rqvy-)p_RYs{%O1*SDk~z}{H6|k)
zBcrP@+~+X)rbVxyUI~vRUR5LMz+O77j%Vw}-bMHQvS8mZ#P>uAB^xdN3@@IFV5{>9
z5~xzHyML?@E2Uwv@H69Pha5K!Pk84tyL7HWeN!lH*gB`wts%{hr;SiDpJuH)NSCOH
z8@)^wd*@vDb%W7#*+g}`i$)ZwjT_pmZ0c_SH|Q8N9mUYxzQJ=kO-mxaB%!tcFrmuR
z^AZUPF5Pwau3!m3mfCJdI2Y?PVVX+#*~kn&b&w4cJWiS?@3ZC-tlH2Sws+5_?f0*c
z(QxC&D~N&*7~RQcxF8hWjM-YYacElDLd*_t*WPNiF+ZhoS>^BP>{f8>s<Te}w3q6j
zE8oZGRdCZ@@}s@l2_*ww_25UV-^!ir4_<zIrUag=9}JTGD}~Ko`G=!ZkhzpRTl4_t
zz<H5c8)0Z_!1vspLyPg0@BMd4jX&z?RUh^fAX{rR(;<t`RCZvxR_*H!pKfLIZ^Nmt
zKL8lgkDV)8qcbjImC@-T(AUJctI!jL_-1#!##b+BBM|I}5<*O(7h&A%w-0~8Du}%L
zA@ZNpsnuhF8_A-9Qa>t8AG9~8n`}p~frxBbplhjNR&kzk#q}#1%@MRIn4qIjrj{OB
z*JM<WigoFWeAY<F5i1JqGsdDl<v@`Y9<mf6A!>T(^DXp*{IOw*Z8ZCo<P=&w?el)g
zrN33M4){D-E@H&SeoCQU2o(3FvMRl@kpn+@o6W$JcqEuY{}hJ>guE{!Ui;}7<7_V*
z>zZzrt!$c|y!So7Y&WD_#1TwKFM&6cS%4f+ILE26c%#>vCi4_a=$F=P7$f*qA)>z<
zFXs*bRSSPL@sIun)H<;UVd^r?z4H}bO1X92`6rCuCY}ELSx$xOgYLtho((l$Wt~Sn
zDFj5YiUodg{%K>%Gif^JZvy$F=iGan0I2g&9$WwN9+<3*?P{#vNzT<tyZ{QZ)leDn
zQ<-{$x5<7Arx|O%mx-s}>CM(IYLEFRNd_T=i5yoKU)Y0Y5;YycuIyjPcT@W3#=xx}
z?SRhS3YQmaXg5O8p|(jH1ube5yYqlE<dz|XDTba?D$_{YgpBJblHt1~wO_o-oAS;U
z7_AqG^uf9#8JV*@vxLig4%lRfsYxwD^$43sLf@H}Audk7a6Xb61r2Rc^74h?NW^nA
zNv3a=OP#8cd8~uhd+guJ5<sD#zsN`Nzfpx~AaD~As1EDJeS<7jz>)elK_<@*G4n8=
zqm-YfO@&e<`eZpV8%0lP<)h42J#h)`?@i}(_tnKTzcJr4<u^-kc#mH4-k=X^;smQ(
z7;Qy`B?K>9NuT5BZzY@R#l`c2qC6zKc}vDJ6T<}S7&8!`{b_o5_3%iFhup6^VH5z#
zrM7aF%(X7h%m}nZW$2LJXbdkTnTifiE`B=K+VHULp1|2HsX1C66^|k#Di{zt%5EBB
zq|X}IHS7MKKoe$w1TX-!GWM>VZFBaTLf4Jzi(@`R_OQ>-gOE#?0B|{3y*5PLtGsfS
zc&@D5I?}&$LFFi3Pi6gpv)TH}b(Q874Lqa|$_XR9x3LP|=BYl^BhW+2);^pGg-a)4
z)HnX>N%log|N89dX2xTc=Jo^Q*$;!ju?MQ;u3PTtqEaL461NQ_!;c7Ry=7b4-%}%b
z;)9^^thP5U9gG}M&3kq(E0j|LQ961TRNy~ULqEL>mOCoMC%==UM2+y(e~HQZ`e3$%
z4_Ny%<+5fk-FAW3V332ncm8hz*NW*3*jK5uHn*yUjqY!Gs^ZKy+9ZE2J+fp!7OOWk
zY%TSRo3&))`dEzG-n2R=R#ljQ&;fi87Qhww8wcDffR9XGSDv#LKhWx5QUZ`_a<E>&
z(5O~{$51$nw&B;uTSU?M?Db7Wz}f`6auc7bZfi6Xs;2u1=Xyy6?C~>9vzbqGmUOch
z(1rp%Y@y`iE7Z@kG#qJ_HmyKDMl^c>%h53QXk*3mCyBu)5#$|H%_yT=k7$r;j0t2a
znEc5;yLJ1vO*p!7+n}k(jA+l4%=!kpdUf?4q|HRJM6{1N-QKk3kT>d~xUgm^;e7;X
z2P{C_7EW0vq^lv!p#vgm5x5IN#h(N;ywEQRZlk>sbZAO_b`g5j2l|l?Xf4DxZYxe`
zqKnA|?y^yUg@(unfXo$rC5o<Tr%mRHRaJ3KLKn*_wI6aqew~}X$!C0JzGk++w*4uj
z(Yfh+a71J+7ZbrPhRU4lFO<q%H)S9lwna~CxCO4>1F2~fYd!q&D{zpls!HovO1@_}
z_#&TRA>b`rek2l;-XUBv?%g+~OipX;3&+V{FNA96rgrB+^d5mtsSk_dw2nPY5Do)o
zXUU>&H&}Sdc=AGpB09^17G6LDli^O$uHs-ge_+;rnI&Kd;27}SylPmow^feo+i^XF
z2E1u%hN(T~8WRIh_gd_~?21qQ5cV{Sn()2t7D+Oy&4VAGyR~vJ;y;?^EtL1tv!qBp
z7ga8>vO(t~jj8}>@PraZ8U0Rgp`-%PrV$X2E;|LsgdEgJE|S}d?YhwPmJAXTDOvpH
zWh3hbbumer(Wx)n6)W&F6-YgBo-$0bll3yp<KXQZ&KYJPK;Er^z4`S)<FRSQzz&!1
zw9mfI&DgLWKV;LqsQcRAe+HkMqDvP60k#&q`Bt33{aGFtbyb^+B@t#-<4dA+&GAfK
z=>N7R06q#}i}#93cs%?twp^hX9mb(4i>1AqWQlgpD<iP_PA3t~w+A(Ryo_|JYSglS
zCUN{qW$@CgD*fmS{TtQv;D>$Met&afsl5M|^9hXBWg8dG2GSZB;R<Y;XU}QI?t@?W
zfljHaC3NEaOu(4O4H&L_<8I0TRjO{NLV0;|#<zK13yomL?b$=|80+qby<*14THMB2
zy{2#Z)l<xGkMIO{|Mdb^p7N^Yq6fQdeIN8b-&DZ6s@+;Mrm9sJ&ot6oB=wg;Ob-O$
zrp%YSu_eQ*L+?JNeWvg{cQ;>tW>S=DCFsyrc{MNRX<1u%C!aemb^;%Ky>CuGZR0RV
z3g>+}Tjl{u+WiTi^^%0e7*hD!g_4w3S2TEu#(5R~8tex|wUo5F(7gMK?a?#^GQf3p
znrQzG%^b@eCRfrC{^@~o7m{zBR30$z9(umCcgvCr+@QfaP3GmE-y3L~wBo=z)I+6G
z5(R5AU{oQa`B1-OwP;$maNt)Yn+=({AH=2l7d=bA*~TaBJ)=^7Y01Ne&1yLAdhT8}
zu7e~$8xKSgZUc=5F3^Aryp-(0bWtnh)PLeT8#qffxBaA=-1IlE(#o1WH12H!8D^8(
z@+1DRX~e-R&JU8OvJ{{}WmP4UJ#rQT2a|`d9zQjAY#y1dxBx+z6L}f=pPJCAnIXFP
zWXr{-stelUyf@aY#kKWNoPoU)(>!JKfZ0lgz2!4d^#iO8XDe6Lu2UNq`h0)$$+RE}
z@{lXo|EGdcL_uYreLjyHM_VxL%sT-nKEze5RYbsJ*ORZ8XyTH0S4jK!`0VO6mB-lv
zcX1iv!+<$NgBdQNQQH}%r(o)h<z0(sAbM)0{pE2u5we~(?%SBcYm_(Z36Eyv7*EyR
zo(RMT%F{yrnB{~tW%2k&Fz<hIumY@hy<slcY<gZuwE)_tXB9%2z<uiDuX;SX<d6or
zdc*a0oQyVT8jS4jA1+zpIt3gwikV$f0dL$#C1Qd#lW+C<`D+LEx0HI~hX6yWmgesZ
zKHwO%EwhzS#duM`TNe0K`m5|`M2Wl)FX5fu5Fp3wCyn^RcSN%5@gIxxmo5CM>b5|1
z`on&}&|>xV(`wC>I`&*Yx!gW6jX$Y)dNKzX^CMgO_jt8qAMa8cJ-O9_%o~Z?^G;{A
zsH)Zw8Sjr&A82JYgpvhLk?hyI{2GUlJ(_6#V3d`Zkq<4cA@rjKL93+HoHlCQk0sxJ
zs5?p8NzimmS+a+0D;|5Q-_WWsp3gg7vKZhS?%6A2MPYVfO?UArRP@$Qz7uY%%-F~|
z_r81qBtFVj#B`>k_DVh84KL|d+}4oTD`Mlk;##5FIc(|kr09Vn){4uL&3C7#x*}v2
z!P7CO%9%mIv#{U+Qbj}A7EQjEh-rb!TED~HD*8#GiNxo?j^iFfy8e}0-k=wwk8Rgw
zn*TkdFlE9{q0E6dj2jI`r&)virA`3L?Pj6e$)#mPa(;yTPkdCzl~UFg@WJJKi^18&
z^NOG8uHz<|J*Z(GG(R@vytazqMkBU@%tp_XzPmR)uhvfdydq_DdLF&^4bO&@Lw8S`
zuwE$>?v};=kzd2{Chm7WiS2{i=Jg0qu+GR8uAfA;BR#h79`Knep9b*0|B0+Z#*C-H
zl|0KiALr6l%<Jq+b;|?SbJ(*Gu0Sr>u6Q>ckww2*wTIV{T|V$Rv}xYo<tC=rsvY^2
zjBArvZ_bs@INY#_sKTm~Ic(KY>5Q-)GJvlH<{ao9?^?geCxzFgJQ9b%WVu&1ZXfet
zx)G^cPNpt8h9DP;8MJmv=FJ|}MkFctAs+~#%8<2W`wS-nG0yxt{iLP)7Sa|7(;YcR
zy#pbW=Jb(6ldzcvLEwT7ONjeCtQw3UwR>NriD&!Om!NFrDCcvzH3b0KfNGwzM{@lp
zrrJfLocdPV?k-x~M8M(0i;}6ZO<JqHb5ac_hp(M=uyQ-6$gYb3{cEhqLf0vztyUFJ
zJh#26&&j8_OD2IyK1CMmug?Ep%Ai*6y@6#wPgTR%#-bdjB2tvj!>MWh1)5~5C!SH$
zw~P(raa*SqmztjiTPZu0@>h%I&Fi$f5boVe+FQn>L1oB(8Zlkd#IP-Ev^=+&deRY}
zx^gf6PkZP7&t&@t@X}}|va&hESMf@VO(o>~v^+T^%yG`CRGTNKgq*6WA(F9B4q1wk
zksOCSeOnU3JUKklnzlwOLgujW<$JB3*X#Qie1Gfy?Q`Gvbzj%#x<1!^U$6Ufecqp%
zUY~9o-9Tu0FdfzmH*F|Lry(Fj%)JhU(+c>PO|Ik~bwlOR-UJVGj}4Y9ug1H5{TpVz
z0_K`^#^t9)uly#1{_YXI25;Su3d~8fme7)lzdRpy7FJsR=0+9Q>U>6oa74AMM={Iu
zabZ<EBOwvU@-Cm9sm@3vre4nHnJ=gAk**XtzKuF@Zj7FtHg!f*aJ~T1IuRa~{o<=&
zpC6=IvXP2{71XN_%O@@AIR`g*rRUt8={jZ=W~2<w*Lh%ukZrl1x9q!USQl1StOgeA
z*OQPfrMcBbSA&O38y&RGS6-N{IMq97<=?y;#SO;xbq2fL3(G62jKstER67m+fJO-w
z(=zzDb;iwp=-wB@#L$e|M38^;(zmAjS$cqx2~mZhaR=X{ncly>9om(i>&^%@IxEYk
zx>)+#wi`|Q2A%+M@MZRow#C16J5WrL_Q{XMVTs@6z>2#dw59b{^>(zGOSx>RaZ9gb
z>vsnvwi^AzS7THadf!)SX3%-#P-~^n&Aal|Fl>UsqpretL4B1=?s9x8-lcZXRtuyf
z2Y!3ca|gdEYXQcxn$M<xwI)}C=w-&k`a$b>@&BGE7xI~A9(Yn$?0|qFv^1t~#7=Rt
z0o=opu&>ta>na(<7-ytM-yr-YZO#3&udaExL^mwT2*lCuT<EuqRx6&TR00eZM77;I
z{$YJRDYUaeDW_FW`pV-Y_dEPeb6IkmcCS?OSL9)t%Ij)rH}_~ATMSsv1ucx-9ow4{
z`)=4>bFE~NKwFG1#6;xEBFkv5`0vAIRNx!@O0e{|7-rG5lWg=kgW+1jRGI@);Sj&C
zF&tZa_(cY@ZsF!mHyY1|Gp71|oFX`!!<59d-<t18@QyhSoWmBr5knN+{@rp1D^_*G
za%ix2B#gSr$d6|q!C8jeq9gHxZyG+ZtLldwKH6UWsYo+)M@XfC(Pj6VKfFJ4seG|g
zqSO60rGQ*1#C)tKgguq?gZkmfW?8kcry2+hnY_(oikUYLHq7<Wwe!X<1x<q-!koDC
z+%~#0Hzv-7#$}Is5>H!Y?!q;fo}KtU4m9g^@;z~~3hqwo#tgAL0!sD^Qy26n;<3O%
zA9vuSV?P|rzU2iKRn*9+!sm~@8C_vRMgNI2nH%8QB<%gRjf`D_&o7eYw_e|e6l@M=
zvbNSXv7zgBKxHYn*hz}6Vj*8u`_`TNd(SU40-Qj)C?>;#A5JdLQ&(x8O)9~Sdzcbd
zk>>@v68wuki*V@iUX#oCk=0hGzAL(C-!iUyJT7A-T1NTPPTeNvJV!n+P;+`(^e0vl
zcJDFE0wMa`E)EKyCLW@jS#EpFrM`w{-k~<wgX;sEWNY-5-AQ*?Jy|g4)V@5Km3YBg
zb!23nF(4>_zR%*`S*o?wwcu>mn0Kb1EMdl)`U1U2$5g!qD1<x7#?4EevEMIWe7VmP
zpdg|f$eT3|9Lrk9PTP@v3!IT74om^E<){t#h|uhW{qM?8MSRR7)5jb+%XZ>7fXcl&
z0EH8|KFjt(AV;-u#oplbHt?VT=aKz^S#pmK<HX@&35}sGt7exlRhJ$ufET9+7%4Io
z4KMJt;SMQ!g{p07q7-aY%!pJ(O(Mi{N5h47){jmP=HywkkI<3}z7*cBI;5#lQS)LX
ztgIIxEn2(sPz<*iTmWDv_}qT-DffX0X!QsmBZFDBSn7n50gR`3)c?!u2YjG59>9K}
z7vKs_dR28(MeUd-&Go~xoUiyKl=GA-<&7)JO3&odE7^$@Zt`|Fp}Z@zQK0ga19%XK
zA?n*t5=u#dp7U6Iumn6oH+PNmO9U+_OB-0a=Lbe*Fl$1rZ<|Sj)z=x%tqr70=qBAt
zi@*>z?_je{M3PeRlMl?N-UE<?SNxncuSZ#C3%$yN7t*K?@Cd*A&RyeaN_ZDlW|w$x
ztLIa+ZG1Wsy`xKBj)Jj%nSMbtLxpJ6_*589E!8N@%=dIN953+~UZGI<i_G0RQ3XEk
zn90uRs;;RDn0skGMmX~Xcm<ERTmu!K$wqea>NZ^sB$WqH)KRD{W?DUw0{HXS(i(7;
z<{H|w%S_g`2r9VKG|}h<0(s(5IP*j^={;L5TeF-8;~hV<;4U#F&K}u)-oP7=Exr~k
z&^WEfz+XJKpy2Ce*3U4QY!f3!Iw1hAmjpdv&Zs$MH&bna%bAT`%mR2FuJe0%xlUB(
zwP>>xp_hL*OD-OaRIulg3}#DM_bMu}7*y^tXI%M82f#Z*`qoTNKV-qY|LA9>mN7DI
z&}hXW7yWb^Old4YOq|#VOe=7sQvLom`YGb@E+EY<U(58`S-o!YNy4uozCR<gXLq^O
z^W#>pt<Yd_$kTH)@=<;=+ZJsxmistTMcRX=7Q2#G_yE{s5wT}*ME*m<tOJFjw{>L8
z--_QttfBepD62527nCnb4j-43CR)GzB#)&SyC-lJAw;Qf_E3Y367R(G2&TVU<&!a*
za8ThQp`Em@u(V`cVLwosBOXY$QmMwB9F|5ZV9u>`ks#eNEN!n6NH3{?N73s!sYwaz
z&tzIng<(S#=;hF&i-eqYQpBAjRYQZ@aVGT~LK6gY{qDRlC~5WIjzzsn^XZT+#;710
zB|+rY!`e!VUxL;QDKx*+2bh*v7>^K1COv?JvTk%7-S8vJdX(f}*iVv{L58HMKZQ<1
z$<~2?gaIg+6aNUuwSN`()l0uj<NrJ#Wb)V66%XPH2yeIs8^B?I)am!?BY{c(1t}{+
AHvj+t

literal 0
HcmV?d00001

diff --git a/index.html b/index.html
index 6ec4e77..f64084d 100644
--- a/index.html
+++ b/index.html
@@ -350,6 +350,43 @@
 </div>
 
 
+<div class="row" id="Machine Learning">
+  <div class="col-sm-12">
+    <h2 class="homepage_category_title">Uncertainty quantification</h2>
+  </div>
+
+
+  <div class="col-sm-12">
+    <div class="panel panel-default">
+      <div class="panel-heading">
+        <h3 class="panel-title">Basics</h3>
+      </div>
+      <div class="panel-body">
+        <ul>
+            <li>Coming soon.</li>
+        </ul>
+      </div>
+    </div>
+  </div>
+  
+    <div class="col-sm-12">
+    <div class="panel panel-default">
+      <div class="panel-heading">
+        <h3 class="panel-title">OpenTurns</h3>
+      </div>
+      <div class="panel-body">
+        <ul>
+            <li>Coming soon.</li>
+        </ul>
+      </div>
+    </div>
+  </div>
+
+
+
+</div>
+
+
 <div class="row" id="cryptocurrencies">
   <div class="col-sm-12">
     <h2 class="homepage_category_title">Cryptocurrencies</h2>
@@ -445,7 +482,7 @@
     <div class="panel panel-default">
       <div class="panel-body">
         <ul>
-            <li>Coming soon.</li>
+            <li><a href="./sql/display_table.html">Displaying the content of a TABLE</a></li>
         </ul>
       </div>
     </div>
@@ -486,7 +523,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/linux/Reloading_.bashrc.html b/linux/Reloading_.bashrc.html
index 3c6f64a..0c8c026 100644
--- a/linux/Reloading_.bashrc.html
+++ b/linux/Reloading_.bashrc.html
@@ -144,7 +144,7 @@ However, you can use the following command to do it </p>
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/linux/Reloading_fstab.html b/linux/Reloading_fstab.html
index b85e4db..a3be230 100644
--- a/linux/Reloading_fstab.html
+++ b/linux/Reloading_fstab.html
@@ -143,7 +143,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/linux/mount_nfs_share_fstab.html b/linux/mount_nfs_share_fstab.html
index f27c6e4..8e4909b 100644
--- a/linux/mount_nfs_share_fstab.html
+++ b/linux/mount_nfs_share_fstab.html
@@ -158,7 +158,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/linux/share_nfs_share.html b/linux/share_nfs_share.html
index 25ced93..600b2eb 100644
--- a/linux/share_nfs_share.html
+++ b/linux/share_nfs_share.html
@@ -151,7 +151,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/machine-learning/Source code for the recommandation engine for articles.html b/machine-learning/Source code for the recommandation engine for articles.html
index 3bc142a..3f9f94c 100644
--- a/machine-learning/Source code for the recommandation engine for articles.html	
+++ b/machine-learning/Source code for the recommandation engine for articles.html	
@@ -1033,7 +1033,7 @@ Top dislike features:
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/pages/about.html b/pages/about.html
index df4230f..5965856 100644
--- a/pages/about.html
+++ b/pages/about.html
@@ -145,7 +145,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Creating_a_sqlite_database.html b/python/Creating_a_sqlite_database.html
index dff604e..9d2cc81 100644
--- a/python/Creating_a_sqlite_database.html
+++ b/python/Creating_a_sqlite_database.html
@@ -193,7 +193,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Iterating_over_a_dataframe.html b/python/Iterating_over_a_dataframe.html
index 8f58cdd..c2c874a 100644
--- a/python/Iterating_over_a_dataframe.html
+++ b/python/Iterating_over_a_dataframe.html
@@ -235,7 +235,7 @@ At line 5 there is a plum which is purple and contains 28 kcal
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Opening_SQLite_database.html b/python/Opening_SQLite_database.html
index c0b2803..e40a9d6 100644
--- a/python/Opening_SQLite_database.html
+++ b/python/Opening_SQLite_database.html
@@ -169,7 +169,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Opening_file.html b/python/Opening_file.html
index 4352f16..b88e143 100644
--- a/python/Opening_file.html
+++ b/python/Opening_file.html
@@ -193,7 +193,7 @@ If you try to open a non existing file in read mode 'r', a FileNotFoundError wil
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Reading_data_from_a_sql_database_with_pandas.html b/python/Reading_data_from_a_sql_database_with_pandas.html
index 33c6cee..41da047 100644
--- a/python/Reading_data_from_a_sql_database_with_pandas.html
+++ b/python/Reading_data_from_a_sql_database_with_pandas.html
@@ -238,7 +238,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Saving_a_matplotlib_figure_with_a_high_resolution.html b/python/Saving_a_matplotlib_figure_with_a_high_resolution.html
index e142f91..687e02d 100644
--- a/python/Saving_a_matplotlib_figure_with_a_high_resolution.html
+++ b/python/Saving_a_matplotlib_figure_with_a_high_resolution.html
@@ -186,7 +186,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html b/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html
index 8de8c28..b4e4c59 100644
--- a/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html
+++ b/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html
@@ -171,7 +171,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/Writing_data_to_a_sql_database_with_pandas.html b/python/Writing_data_to_a_sql_database_with_pandas.html
index 919403b..9c70fb0 100644
--- a/python/Writing_data_to_a_sql_database_with_pandas.html
+++ b/python/Writing_data_to_a_sql_database_with_pandas.html
@@ -340,7 +340,7 @@ However, exporting to a sql database can prove very useful. Indeed, having a wel
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/python/updating_all_python_package_with_anaconda.html b/python/updating_all_python_package_with_anaconda.html
index 1a4d28a..738f137 100644
--- a/python/updating_all_python_package_with_anaconda.html
+++ b/python/updating_all_python_package_with_anaconda.html
@@ -151,7 +151,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/search.html b/search.html
index f898c35..a8727f8 100644
--- a/search.html
+++ b/search.html
@@ -112,7 +112,7 @@
     <footer class="footer">
         <div class="container">
             <p class="text-muted">
-                <center>This project contains 16 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
                 <br/>
                 Copyright &copy; Guillaume Redoulès,
                     <time datetime="2018">2018</time>.
diff --git a/sitemap.xml b/sitemap.xml
index d648ebe..11a4801 100644
--- a/sitemap.xml
+++ b/sitemap.xml
@@ -5,7 +5,7 @@ xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
 
 <url>
 <loc>redoules.github.io/</loc>
-<lastmod>2018-08-03T18:53:34-00:00</lastmod>
+<lastmod>2018-08-04T16:28:24-00:00</lastmod>
 <changefreq>daily</changefreq>
 <priority>0.5</priority>
 </url>
@@ -17,6 +17,13 @@ xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
 <priority>0.5</priority>
 </url>
 
+<url>
+<loc>redoules.github.io/sql/display_table.html</loc>
+<lastmod>2018-08-04T16:14:00+02:00</lastmod>
+<changefreq>monthly</changefreq>
+<priority>0.5</priority>
+</url>
+
 <url>
 <loc>redoules.github.io/python/Opening_file.html</loc>
 <lastmod>2018-08-03T18:25:00+02:00</lastmod>
diff --git a/sql/display_table.html b/sql/display_table.html
new file mode 100644
index 0000000..52f649e
--- /dev/null
+++ b/sql/display_table.html
@@ -0,0 +1,265 @@
+<!DOCTYPE html>
+<html lang="fr">
+
+<head>
+      <meta charset="utf-8">
+    <meta http-equiv="X-UA-Compatible" content="IE=edge">
+    <meta name="viewport" content="width=device-width, initial-scale=1">
+    <!-- The above 3 meta tags *must* come first in the head; any other head content must come *after* these tags -->
+    <meta name="description" content="Data Science for Political and Social Phenomena">
+    <meta name="author" content="Guillaume Redoulès">
+    <link rel="icon" href="../favicon.ico">
+
+    <title>Displaying the content of a TABLE - SQL</title>
+
+    <!-- JQuery -->
+    <script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.3/jquery.min.js"></script>
+    <script>
+        window.jQuery || document.write('<script src="../theme/js/jquery.min.js"><\/script>')
+    </script>
+
+    <!-- Bootstrap core CSS -->
+    <link rel="stylesheet" href="../theme/css/bootstrap.css" />
+    <!-- IE10 viewport hack for Surface/desktop Windows 8 bug -->
+    <link rel="stylesheet" type="text/css" href="../theme/css/ie10-viewport-bug-workaround.css" />
+    <!-- Custom styles for this template -->
+    <link rel="stylesheet" type="text/css" href="../theme/css/style.css" />
+    <link rel="stylesheet" type="text/css" href="../theme/css/notebooks.css" />
+    <link href='https://fonts.googleapis.com/css?family=PT+Serif:400,700|Roboto:400,500,700' rel='stylesheet' type='text/css'>
+
+    <!-- HTML5 shim and Respond.js for IE8 support of HTML5 elements and media queries -->
+    <!--[if lt IE 9]>
+      <script src="https://oss.maxcdn.com/html5shiv/3.7.2/html5shiv.min.js"></script>
+      <script src="https://oss.maxcdn.com/respond/1.4.2/respond.min.js"></script>
+    <![endif]-->
+        
+
+    <meta name="tags" content="Basics" />
+
+
+</head>
+
+<body>
+
+    <div class="navbar navbar-fixed-top">
+        <div class="container">
+            <div class="navbar-header">
+                <button type="button" class="navbar-toggle" data-toggle="collapse" data-target=".navbar-collapse">
+                    <span class="icon-bar"></span>
+                    <span class="icon-bar"></span>
+                    <span class="icon-bar"></span>
+                </button>
+                <a class="navbar-brand" href="..">Guillaume Redoulès</a>
+            </div>
+            <div class="navbar-collapse collapse" id="searchbar">
+
+                <ul class="nav navbar-nav navbar-right">
+                    <li class="dropdown">
+                        <a href="#" class="dropdown-toggle" data-toggle="dropdown" role="button" aria-haspopup="true" aria-expanded="false">About<span class="caret"></span></a>
+                        <ul class="dropdown-menu">
+                            <li><a href="../pages/about.html">About Guillaume</a></li>
+                            <li><a href="https://github.com/redoules">GitHub</a></li>
+                            <li><a href="https://www.linkedin.com/in/guillaume-redoul%C3%A8s-33923860/">LinkedIn</a></li>
+                        </ul>
+                    </li>
+                    <li class="dropdown">
+                        <a href="#" class="dropdown-toggle" data-toggle="dropdown" role="button" aria-haspopup="true" aria-expanded="false">Data Science<span class="caret"></span></a>
+                        <ul class="dropdown-menu">
+                            <li><a href="..#Blog">Blog</a></li>
+                            <li><a href="..#Python">Python</a></li>
+                            <li><a href="..#Bash">Bash</a></li>
+                            <li><a href="..#SQL">SQL</a></li>
+                            <li><a href="..#Mathematics">Mathematics</a></li>
+                            <li><a href="..#Machine_Learning">Machine Learning</a></li>
+                            <li><a href="..#Projects">Projects</a></li>
+                        </ul>
+                    </li>
+                    <li class="dropdown">
+                        <a href="#" class="dropdown-toggle" data-toggle="dropdown" role="button" aria-haspopup="true" aria-expanded="false">Projects<span class="caret"></span></a>
+                        <ul class="dropdown-menu">
+                            <li><a href="https://github.com/redoules/redoules.github.io">Notes (Github)</a></li>
+                        </ul>
+                    </li>
+
+                    <!--<li class="dropdown">
+                        <a href="../feeds/blog.rss.xml">Blog RSS</a>
+                    </li>-->
+
+
+                </ul>
+
+                <form class="navbar-form" action="../search.html" onsubmit="return validateForm(this.elements['q'].value);">
+                    <div class="form-group" style="display:inline;">
+                        <div class="input-group" style="display:table;">
+                            <span class="input-group-addon" style="width:1%;"><span class="glyphicon glyphicon-search"></span></span>
+                            <input class="form-control search-query" name="q" id="tipue_search_input" placeholder="e.g. scikit KNN, pandas merge" required autocomplete="off" type="text">
+                        </div>
+                    </div>
+                </form>
+
+            </div>
+            <!--/.nav-collapse -->
+        </div>
+    </div>
+
+
+
+    <!-- end of header section -->
+    <div class="container">
+<!-- <div class="alert alert-warning" role="alert">
+    Did you find this page useful? Please do me a quick favor and <a href="#" class="alert-link">endorse me for data science on LinkedIn</a>.
+</div> -->
+<section id="content" class="body">
+    <header>
+    <h1>
+      Displaying the content of a TABLE
+    </h1>
+<ol class="breadcrumb">
+    <li>
+        <time class="published" datetime="2018-08-04T16:14:00+02:00">
+            04 août 2018
+        </time>
+    </li>
+    <li>SQL</li>
+    <li>Basics</li>
+</ol>
+</header>
+<div class='article_content'>
+<h1>Displaying the content of a TABLE in SQL</h1>
+<p>In this very simple example we will see how to display the content of a table.
+Since we are working in the notebook, we will load the sql extension in order to manipulate the database.
+The database mydatabase.db is a SQLite database already created before the example.</p>
+<div class="highlight"><pre><span></span><span class="c1">#load the extension</span>
+<span class="o">%</span><span class="n">load_ext</span> <span class="n">sql</span> 
+<span class="c1">#connect to the database</span>
+<span class="o">%</span><span class="n">sql</span> <span class="n">sqlite</span><span class="p">:</span><span class="o">///</span><span class="n">mydatabase</span><span class="o">.</span><span class="n">db</span> 
+</pre></div>
+
+
+<div class="highlight"><pre><span></span>&#39;Connected: @mydatabase.db&#39;
+</pre></div>
+
+
+<p>In order to extract all the values from a table, we will use the following command :
+SELECT * FROM TABLE
+In our example, we want to display the data contained in the table named tutyfrutty</p>
+<div class="highlight"><pre><span></span><span class="o">%</span><span class="n">sql</span> <span class="n">SELECT</span> <span class="o">*</span> <span class="n">FROM</span> <span class="n">tutyfrutty</span>
+</pre></div>
+
+
+<div class="highlight"><pre><span></span> * sqlite:///mydatabase.db
+Done.
+</pre></div>
+
+
+<table>
+    <tr>
+        <th>index</th>
+        <th>fruit</th>
+        <th>color</th>
+        <th>kcal</th>
+    </tr>
+    <tr>
+        <td>0</td>
+        <td>Banana</td>
+        <td>yellow</td>
+        <td>89</td>
+    </tr>
+    <tr>
+        <td>1</td>
+        <td>Orange</td>
+        <td>orange</td>
+        <td>47</td>
+    </tr>
+    <tr>
+        <td>2</td>
+        <td>Apple</td>
+        <td>red</td>
+        <td>52</td>
+    </tr>
+    <tr>
+        <td>3</td>
+        <td>lemon</td>
+        <td>yellow</td>
+        <td>15</td>
+    </tr>
+    <tr>
+        <td>4</td>
+        <td>lime</td>
+        <td>green</td>
+        <td>30</td>
+    </tr>
+    <tr>
+        <td>5</td>
+        <td>plum</td>
+        <td>purple</td>
+        <td>28</td>
+    </tr>
+    <tr>
+        <td>7</td>
+        <td>Cranberry</td>
+        <td>red</td>
+        <td>308</td>
+    </tr>
+</table>
+</div>
+    <aside>
+    <div class="bug-reporting__panel">
+        <h3>Find an error or bug? Have a suggestion?</h3>
+        <p>Everything on this site is avaliable on GitHub. Head on over and <a href='https://github.com/redoules/redoules.github.io/issues/new'>submit an issue.</a> You can also message me directly by <a href='mailto:guillaume.redoules@gadz.org'>email</a>.</p>
+    </div>
+    </aside>
+</section>
+
+    </div>
+    <!-- start of footer section -->
+    <footer class="footer">
+        <div class="container">
+            <p class="text-muted">
+                <center>This project contains 17 pages and is available on <a href="https://github.com/redoules/redoules.github.io">GitHub</a>.
+                <br/>
+                Copyright &copy; Guillaume Redoulès,
+                    <time datetime="2018">2018</time>.
+                </center>
+            </p>
+        </div>
+    </footer>
+
+    <!-- This jQuery line finds any span that contains code highlighting classes and then selects the parent <pre> tag and adds a border. This is done as a workaround to visually distinguish the code inputs and outputs -->
+    <script>
+        $( ".hll, .n, .c, .err, .k, .o, .cm, .cp, .c1, .cs, .gd, .ge, .gr, .gh, .gi, .go, .gp, .gs, .gu, .gt, .kc, .kd, .kn, .kp, .kr, .kt, .m, .s, .na, .nb, .nc, .no, .nd, .ni, .ne, .nf, .nl, .nn, .nt, .nv, .ow, .w, .mf, .mh, .mi, .mo, .sb, .sc, .sd, .s2, .se, .sh, .si, .sx, .sr, .s1, .ss, .bp, .vc, .vg, .vi, .il" ).parent( "pre" ).css( "border", "1px solid #DEDEDE" );
+    </script>
+
+    
+    <!-- Load Google Analytics -->
+    <script>
+        /*
+        (function(i, s, o, g, r, a, m) {
+            i['GoogleAnalyticsObject'] = r;
+            i[r] = i[r] || function() {
+                (i[r].q = i[r].q || []).push(arguments)
+            }, i[r].l = 1 * new Date();
+            a = s.createElement(o),
+                m = s.getElementsByTagName(o)[0];
+            a.async = 1;
+            a.src = g;
+            m.parentNode.insertBefore(a, m)
+        })(window, document, 'script', '//www.google-analytics.com/analytics.js', 'ga');
+
+        ga('create', 'UA-66582-32', 'auto');
+        ga('send', 'pageview');
+        */
+    </script>
+    <!-- End of Google Analytics -->
+
+    <!-- Bootstrap core JavaScript
+      ================================================== -->
+    <!-- Placed at the end of the document so the pages load faster -->
+    <script src="../theme/js/bootstrap.min.js"></script>
+    <!-- IE10 viewport hack for Surface/desktop Windows 8 bug -->
+    <script src="../theme/js/ie10-viewport-bug-workaround.js"></script>
+
+
+</body>
+
+</html>
\ No newline at end of file
diff --git a/tipuesearch_content.json b/tipuesearch_content.json
index bb5d6c1..b0529d7 100644
--- a/tipuesearch_content.json
+++ b/tipuesearch_content.json
@@ -1 +1 @@
-{"pages":[{"title":"About Guillaume Redoulès","text":"I am a data scientist and a mechanical engineer working on numerical methods for stress computations in the field of rocket propulsion. Prior to that, I've got a MSc in Computational Fluid Dynamics and aerodynamics from Imperial College London. Email: guillaume.redoules@gadz.org Linkedin: Guillaume Redoulès Curriculum Vitae Experience Thermomecanical method and tools engineer , Ariane Group , 2015 - Present In charge of tools and methods related to thermomecanical computations. Focal point for machine learning. Education MSc Advanced Computational Methods for Aeronautics, Flow Management and Fluid-Structure Interaction , Imperial College London, London. 2013 Dissertation: \"Estimator design for fluid flows\" Fields: Aeronautics, aerodynamics, computational fluid dynamics, numerical methods Arts et Métiers Paristech , France, 2011 Generalist engineering degree Fields: Mechanics, electrical engineering, casting, machining, project management, finance, IT, etc.","tags":"pages","url":"redoules.github.io/pages/about.html","loc":"redoules.github.io/pages/about.html"},{"title":"Opening a file with python","text":"Opening a file with python This short article show you how to open a file using python. We will use the with keyword in order to avoid having to close the file. There is no need to import anything in order to open a file. All the function related to file manipulation are part of the python standard library In order to open a file, we will use the function open. This function takes two arguments : the path of the file the mode you want to open the file The mode can be : 'r' : read 'w' : write 'a' : append (writes at the end of the file) 'b' : binary mode 'x' : exclusive creation 't' : text mode (by default) Note that if the file does not exit it will be created if you use the following options \"w\", \"a\", \"x\". If you try to open a non existing file in read mode 'r', a FileNotFoundError will be returned. It is possible to combine multiple options together. For instance, you can open a file in binary mode for writing using the 'wb' option. Python distinguishes between binary and text I/O. Files opened in binary mode return contents as bytes objects without any decoding. In text mode , the contents of the file are returned as str, the bytes having been first decoded using a platform-dependent encoding or using the specified encoding if given. Writing to a file Let's first open (create) a text file a write a string to it. filepath = \". \\\\ myfile.txt\" with open ( filepath , 'w' ) as f : f . write ( \"Hello world !\" ) Reading a file we can now see how to read the content of a file. To do so, we will use the 'r' option with open ( filepath , \"r\" ) as f : content = f . read () print ( content ) Hello world ! A word on the with keyword In python the with keyword is used when working with unmanaged resources (like file streams). The python documentation tells us that : The with statement clarifies code that previously would use try...finally blocks to ensure that clean-up code is executed. In this section, I'll discuss the statement as it will commonly be used. In the next section, I'll examine the implementation details and show how to write objects for use with this statement. The with statement is a control-flow structure whose basic structure is: with expression [ as variable ]: with - block The expression is evaluated, and it should result in an object that supports the context management protocol (that is, has enter () and exit () methods).","tags":"Python","url":"redoules.github.io/python/Opening_file.html","loc":"redoules.github.io/python/Opening_file.html"},{"title":"Opening a SQLite database with python","text":"Opening a SQLite database with python This short article show you how to connect to a SQLite database using python. We will use the with keyword in order to avoid having to close the database. In order to connect to the database, we will have to import sqlite3 import sqlite3 from sqlite3 import Error In python the with keyword is used when working with unmanaged resources (like file streams). The python documentation tells us that : The with statement clarifies code that previously would use try...finally blocks to ensure that clean-up code is executed. In this section, I'll discuss the statement as it will commonly be used. In the next section, I'll examine the implementation details and show how to write objects for use with this statement. The with statement is a control-flow structure whose basic structure is: with expression [ as variable ]: with - block The expression is evaluated, and it should result in an object that supports the context management protocol (that is, has enter () and exit () methods). db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : print ( \"Connected to the database\" ) #your code here except Error as e : print ( e ) Connected to the database","tags":"Python","url":"redoules.github.io/python/Opening_SQLite_database.html","loc":"redoules.github.io/python/Opening_SQLite_database.html"},{"title":"Reading data from a sql database with pandas","text":"Reading data from a sql database with pandas When manipulating you data using pandas, it is sometimes useful to pull data from a database. In this tutorial, we will see how to querry a dataframe from a sqlite table. Note than it would also work with any other sql database a long as you change the connxion to the one that suits your needs. First let's import pandas and sqlite3 import pandas as pd import sqlite3 from sqlite3 import Error We want to store the table tutyfrutty in our dataframe. To do so, we will query all the elements present in the tutyfrutty TABLE with the command : SELECT * FROM tutyfrutty db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : df = pd . read_sql ( \"SELECT * FROM tutyfrutty\" , conn ) del df [ \"index\" ] #juste delete the index column that was stored in the table except Error as e : print ( e ) df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 6 Cranberry red 308 7 Cranberry red 308","tags":"Python","url":"redoules.github.io/python/Reading_data_from_a_sql_database_with_pandas.html","loc":"redoules.github.io/python/Reading_data_from_a_sql_database_with_pandas.html"},{"title":"Writing data to a sql database with pandas","text":"Writing data to a sql database with pandas When manipulating you data using pandas, it is sometimes useful to store a dataframe. Pandas provides multiple ways to export dataframes. The most common consist in exporting to a csv, a pickle, to hdf or to excel. However, exporting to a sql database can prove very useful. Indeed, having a well structured database is a great for storing all the data related to your analysis in one place. In this tutorial, we will see how to store a dataframe in a new table of a sqlite dataframe. Note than it would also work with any other sql database a long as you change the connxion to the one that suits your needs. First let's import pandas and sqlite3 import pandas as pd import sqlite3 from sqlite3 import Error # Example dataframe raw_data = { 'fruit' : [ 'Banana' , 'Orange' , 'Apple' , 'lemon' , \"lime\" , \"plum\" ], 'color' : [ 'yellow' , 'orange' , 'red' , 'yellow' , \"green\" , \"purple\" ], 'kcal' : [ 89 , 47 , 52 , 15 , 30 , 28 ] } df = pd . DataFrame ( raw_data , columns = [ 'fruit' , 'color' , 'kcal' ]) df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 Now that the DataFrame has been created, let's push it to the sqlite database called mydatabase.db in a new table called tutyfrutty db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : df . to_sql ( \"tutyfrutty\" , conn ) except Error as e : print ( e ) except ValueError : print ( \"The TABLE tutyfrutty already exists, read below to understand how to handle this case\" ) Note that if the table tutyfrutty was already existing, the to_sql function will return a ValueError. This is where, the if_exists option comes into play. Let's look at the docstring of this function : \"\"\" if_exists : {'fail', 'replace', 'append'}, default 'fail' - fail: If table exists, do nothing. - replace: If table exists, drop it, recreate it, and insert data. - append: If table exists, insert data. Create if does not exist. \"\"\" Let's say, I want to update my dataframe with some new rows df . loc [ len ( df ) + 1 ] = [ 'Cranberry' , 'red' , 308 ] df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 7 Cranberry red 308 8 Cranberry red 308 I can now replace the table with the new values using the \"replace\" option db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : df . to_sql ( \"tutyfrutty\" , conn , if_exists = \"replace\" ) except Error as e : print ( e )","tags":"Python","url":"redoules.github.io/python/Writing_data_to_a_sql_database_with_pandas.html","loc":"redoules.github.io/python/Writing_data_to_a_sql_database_with_pandas.html"},{"title":"Creating a sqlite database","text":"Creating a sqlite database When you want to start with using databases SQlite is a great tool. It provides an easy onramp to learn and prototype you database with a SQL compatible database. First, let's import the libraries we need import sqlite3 from sqlite3 import Error SQlite doesn't need a database server, however, you have to start by creating an empty database file import os def check_for_db_file (): if os . path . exists ( \"mydatabase.db\" ): print ( \"the database is ready\" ) else : print ( \"no database found\" ) check_for_db_file () no database found Let's then create a function that will connect to a database, print the verison of sqlite and then close the connexion to the database. def create_database ( db_file ): \"\"\" create a database connection to a SQLite database \"\"\" try : with sqlite3 . connect ( db_file ) as conn : print ( \"database created with sqlite3 version {0}\" . format ( sqlite3 . version )) except Error as e : print ( e ) create_database ( \".\\mydatabase.db\" ) database created with sqlite3 version 2.6.0 check_for_db_file () the database is ready You're all set. From now on, you can open the database and write sql querries into it.","tags":"Python","url":"redoules.github.io/python/Creating_a_sqlite_database.html","loc":"redoules.github.io/python/Creating_a_sqlite_database.html"},{"title":"Setting up the notebook for plotting with matplotlib","text":"Importing Matplotlib First we need to import pyplot, a collection of command style functions that make matplotlib work like MATLAB. Let's, as well, use the magic command %matplotlib inline in order to display the figures in the notebook import matplotlib.pyplot as plt % matplotlib inline # this doubles image size, but we'll do it manually below # %config InlineBackend.figure_format = 'retina' The following parameters are recommended for matplotlib, they will make matplotlib output a better quality image # %load snippets/matplot_setup.py plt . rcParams [ 'savefig.dpi' ] = 300 plt . rcParams [ 'figure.dpi' ] = 163 plt . rcParams [ 'figure.autolayout' ] = False plt . rcParams [ 'figure.figsize' ] = 20 , 12 plt . rcParams [ 'axes.labelsize' ] = 18 plt . rcParams [ 'axes.titlesize' ] = 20 plt . rcParams [ 'font.size' ] = 16 plt . rcParams [ 'lines.linewidth' ] = 2.0 plt . rcParams [ 'lines.markersize' ] = 8 plt . rcParams [ 'legend.fontsize' ] = 14 plt . rcParams [ 'text.usetex' ] = False # True activates latex output in fonts! plt . rcParams [ 'font.family' ] = \"serif\" plt . rcParams [ 'font.serif' ] = \"cm\" plt . rcParams [ 'text.latex.preamble' ] = \" \\\\ usepackage{subdepth}, \\\\ usepackage{type1cm}\" You can change the second line in order to fit your display. 163 dpi corresponds to a Dell Ultra HD 4k P2715Q. You can check your screen's dpi count at http://dpi.lv/","tags":"Python","url":"redoules.github.io/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html","loc":"redoules.github.io/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html"},{"title":"Why using a blockchain is a bad idea for your business","text":"What having a blockchain implies? storage costs : everyone maintaining the ledger needs to store every transaction bandwith costs : everyone has to broadcast every transaction computational costs : every node has to validate the blockchain control : the creator does not control the blockchain, everyone collectively controls it developpement costs : developping on a blockchain is way harder than on a traditionnal database What to ask a business when they tell you that they are using a blockchain? When a business is telling you about their innovative technology leveraging the power of the blockchain this should immedialty spake some questions : What is the consensus algorithm? who is responsible for validating the consensus rules? what is the nature of the participation ? is it open to access? is it open to innovation? is it a public ledger? is it transparent? does it improves acountability? is it cross borders? how is it validated?","tags":"Cryptocurrencies","url":"redoules.github.io/cryptocurrencies/blockchain_bad.html","loc":"redoules.github.io/cryptocurrencies/blockchain_bad.html"},{"title":"Synology NFS share","text":"Setting up a NFS share login to your DSM admin account, open the \"Control Panel\" and go to \"File Services\" Make sure NFS is enabled Back in the control panel, go to \"Shared Folder\" Select the folder you want to share and clic \"Edit\" Go to the \"NFS Permissions tab and clic \"Create\", add the ip of the device you want to mount the mapped drive on. Make sure you copy the \"Mount path\"","tags":"Linux","url":"redoules.github.io/linux/share_nfs_share.html","loc":"redoules.github.io/linux/share_nfs_share.html"},{"title":"Mount a NFS share using fstab","text":"Mount nfs using fstab The fstab file, generally located at /etc/fstab lists the differents partitions and where to load them on the filesystem. You can edit this file as root by using the following command sudo nano /etc/fstab in the following example, we want to mount a NFS v3 share from : server : 192.168.1.2 mountpoint (on the server) : /volumeUSB2/usbshare * mountlocation (on the client) : /mnt we specify 192 .168.1.2:/volumeUSB2/usbshare /mnt nfs nfsvers = 3 ,users 0 0 the client will then automatically mount the share ont /mnt at startup. Related you can reload the fstab file using this method : https://redoules.github.io/linux/Reloading_fstab.html You can create a NFS share on a Synology using the method : https://redoules.github.io/linux/share_nfs_share.html","tags":"Linux","url":"redoules.github.io/linux/mount_nfs_share_fstab.html","loc":"redoules.github.io/linux/mount_nfs_share_fstab.html"},{"title":"Installing bitcoind on raspberry pi","text":"Installing bitcoind on linux Running a full bitcoin node helps the bitcoin network to accept, validate and relay transactions. If you want to volunteer some spare computing and bandwidth resources to run a full node and allow Bitcoin to continue to grow you can grab an inexpensive and power efficient raspberry pi and turn it into a full node. There are plenty of tutorials on the Internet explaining how to install a bitcoin full node; this tutorial won't go over setting up a raspberry pi and using ssh. In order to store the full blockchain we will mount a network drive and tell bitcoind to use this mapped drive as the data directory. Download the bitcoin client Go to https://bitcoin.org/en/download Copy the URL for the ARM 32 bit version and download it onto your raspberry pi. wget https://bitcoin.org/bin/bitcoin-core-0.15.1/bitcoin-0.15.1-arm-linux-gnueabihf.tar.gz Locate the downloaded file and extract it using the arguement xzf tar xzf bitcoin-0.15.1-arm-linux-gnueabihf.tar.gz a new directory bitcoin-0.15.1 will be created, it contrains the files we need to install the software Install the bitcoin client We will install the content by copying the binaries located in the bin folder into /usr/local/bin by using the install command. You must use sudo because it will write data to a system directory sudo install -m 0755 -o root -g root -t /usr/local/bin bitcoin-0.15.1/bin/* Launch the bitcoin core client by running bitcoind -daemon Configuration of the node Start your node at boot Starting you node automatically at boot time is a good idea because it doesn't require a manual action from the user. The simplest way to achive this is to create a cronjob. Run the following command crontab -e Select the text editor of your choice, then add the following line at the end of the file @reboot bitcoind -daemon Save the file and exit; the updated crontab file will be installed for you. Full Node If you can afford to download and store all the blockchain, you can run a full node. At the time of writing, the blockchain is 150Go ( https://blockchain.info/fr/charts/blocks-size ). Tree ways to store this are : use a microSD with 256Go or more add a thumbdrive or an external drive to your raspberry pi * mount a network drive from a NAS If you have purchased a big SD card then you can leave the default location for the blockchain data (~/.bitcoin/). Otherwise, you will have to change the datadir location to where your drive is mounted (in my case I have mounted it to /mnt) In order to configure your bitcoin client, edit/create the file bitcoin.conf located in ~/.bitcoin/ nano ~/.bitcoin/bitcoin.conf copy the following text # From redoules.github.io # This config should be placed in following path: # ~/.bitcoin/bitcoin.conf # [core] # Specify a non-default location to store blockchain and other data. datadir=/mnt # Set database cache size in megabytes; machines sync faster with a larger cache. Recommend setting as high as possible based upon mach$ dbcache=100 # Keep at most <n> unconnectable transactions in memory. maxorphantx=10 # Keep the transaction memory pool below <n> megabytess. maxmempool=50 # [network] # Maintain at most N connections to peers. maxconnections=40 # Tries to keep outbound traffic under the given target (in MiB per 24h), 0 = no limit. maxuploadtarget=5000 Check https://jlopp.github.io/bitcoin-core-config-generator it is a handy site to edit the bitcoin.conf file Pruning node If you don't want to store the entire blockchain you can run a pruning node which reduces storage requirements by enabling pruning (deleting) of old blocks. Let's say you want to allocated at most 5Go to the blockchain, then specify prune=5000 into your bitcoin.conf file. Edit/create the file bitcoin.conf located in ~/.bitcoin/ nano ~/.bitcoin/bitcoin.conf copy the following text # From redoules.github.io # This config should be placed in following path: # ~/.bitcoin/bitcoin.conf # [core] # Set database cache size in megabytes; machines sync faster with a larger cache. Recommend setting as high as possible based upon mach$ dbcache=100 # Keep at most <n> unconnectable transactions in memory. maxorphantx=10 # Keep the transaction memory pool below <n> megabytess. maxmempool=50 # Reduce storage requirements by only storing most recent N MiB of block. This mode is incompatible with -txindex and -rescan. WARNING: Reverting this setting requires re-downloading the entire blockchain. (default: 0 = disable pruning blocks, 1 = allow manual pruning via RPC, greater than 550 = automatically prune blocks to stay under target size in MiB). prune=5000 # [network] # Maintain at most N connections to peers. maxconnections=40 # Tries to keep outbound traffic under the given target (in MiB per 24h), 0 = no limit. maxuploadtarget=5000 Checking if your node is public one of the best way to help the bitcoin network is to allow your node to be visible and to propagate block to other nodes. The bitcoin protocole uses port 8333, other clients should be able to share information with your client. Run ifconfig and check if you have an ipv6 adresse (look for adr inet6:) IPV6 Get the global ipv6 adresse of your raspberry pi Link encap:Ethernet HWaddr xx:xx:xx:xx:xx:xx inet adr:192.168.1.x Bcast:192.168.1.255 Masque:255.255.255.0 adr inet6: xxxx::xxxx:xxxx:xxxx:xxxx/64 Scope:Lien adr inet6: xxxx:xxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx/64 Scope:Global UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:42681744 errors:0 dropped:0 overruns:0 frame:0 TX packets:38447218 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 lg file transmission:1000 RX bytes:3044414780 (2.8 GiB) TX bytes:2599878680 (2.4 GiB) it is located between adr inet4 and Scope:Global adr inet6: xxxx:xxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx/64 Scope:Global Copy this adresse and past it into the search field on https://bitnodes.earn.com/ If your node is visible, it will appear on the website IPV4 If you don't have an ipv6 adresse, you will have to open port 8333 on your router and redirect it to the internal IP of your raspberry pi. It is not detailed here because the configuration depends on your router.","tags":"Cryptocurrencies","url":"redoules.github.io/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html","loc":"redoules.github.io/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html"},{"title":"Reloading .bashrc","text":"Reload .bashrc The .bashrc file, located at ~/.bashrc allows a user to personalize its bash shell. If you edit this file, the changes won't be loaded without login out and back in. However, you can use the following command to do it source ~/.bashrc","tags":"Linux","url":"redoules.github.io/linux/Reloading_.bashrc.html","loc":"redoules.github.io/linux/Reloading_.bashrc.html"},{"title":"Reloading fstab","text":"Reload fstab The fstab file, generally located at /etc/fstab lists the differents partitions and where to load them on the filesystem. If you edit this file, the changes won't be automounted. You either have to reboot your system of use the following command as root mount -a","tags":"Linux","url":"redoules.github.io/linux/Reloading_fstab.html","loc":"redoules.github.io/linux/Reloading_fstab.html"},{"title":"Updating all python package with anaconda","text":"Updating anaconda packages All packages managed by conda can be updated with the following command : conda update --all Updating other packages with pip For the other packages, the pip package manager can be used. Unfortunately pip hasn't the same update all fonctionnality. import pip from subprocess import call for dist in pip . get_installed_distributions (): print ( \"updating {0}\" . format ( dist )) call ( \"pip install --upgrade \" + dist . project_name , shell = True )","tags":"Python","url":"redoules.github.io/python/updating_all_python_package_with_anaconda.html","loc":"redoules.github.io/python/updating_all_python_package_with_anaconda.html"},{"title":"Saving a matplotlib figure with a high resolution","text":"creating a matplotlib figure #Importing matplotlib % matplotlib inline import matplotlib.pyplot as plt import numpy as np Drawing a figure # Fixing random state for reproducibility np . random . seed ( 19680801 ) mu , sigma = 100 , 15 x = mu + sigma * np . random . randn ( 10000 ) # the histogram of the data n , bins , patches = plt . hist ( x , 50 , normed = 1 , facecolor = 'g' , alpha = 0.75 ) plt . xlabel ( 'Smarts' ) plt . ylabel ( 'Probability' ) plt . title ( 'Histogram of IQ' ) plt . text ( 60 , . 025 , r '$\\mu=100,\\ \\sigma=15$' ) plt . axis ([ 40 , 160 , 0 , 0.03 ]) plt . grid ( True ) plt . show () Saving the figure normally, one would use the following code plt . savefig ( 'filename.png' ) <matplotlib.figure.Figure at 0x2e45e92f400> The figure in then exported to the file \"filename.png\" with a standard resolution. In adittion, you can specify the dpi arg to some scalar value, for example: plt . savefig ( 'filename_hi_dpi.png' , dpi = 300 ) <matplotlib.figure.Figure at 0x2e462164898>","tags":"Python","url":"redoules.github.io/python/Saving_a_matplotlib_figure_with_a_high_resolution.html","loc":"redoules.github.io/python/Saving_a_matplotlib_figure_with_a_high_resolution.html"},{"title":"Iterating over a DataFrame","text":"Create a sample dataframe # Import modules import pandas as pd # Example dataframe raw_data = { 'fruit' : [ 'Banana' , 'Orange' , 'Apple' , 'lemon' , \"lime\" , \"plum\" ], 'color' : [ 'yellow' , 'orange' , 'red' , 'yellow' , \"green\" , \"purple\" ], 'kcal' : [ 89 , 47 , 52 , 15 , 30 , 28 ] } df = pd . DataFrame ( raw_data , columns = [ 'fruit' , 'color' , 'kcal' ]) df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 Using the iterrows method Pandas DataFrames can return a generator with the iterrrows method. It can then be used to loop over the rows of the DataFrame for index , row in df . iterrows (): print ( \"At line {0} there is a {1} which is {2} and contains {3} kcal\" . format ( index , row [ \"fruit\" ], row [ \"color\" ], row [ \"kcal\" ])) At line 0 there is a Banana which is yellow and contains 89 kcal At line 1 there is a Orange which is orange and contains 47 kcal At line 2 there is a Apple which is red and contains 52 kcal At line 3 there is a lemon which is yellow and contains 15 kcal At line 4 there is a lime which is green and contains 30 kcal At line 5 there is a plum which is purple and contains 28 kcal","tags":"Python","url":"redoules.github.io/python/Iterating_over_a_dataframe.html","loc":"redoules.github.io/python/Iterating_over_a_dataframe.html"},{"title":"Article Recommander","text":"import pandas as pd import numpy as np % matplotlib inline Loading data and preprocessing we first learn the pickled article database. We will be cleaning it and separating the interesting articles from the uninteresting ones. df = pd . read_pickle ( './article.pkl' ) del df [ \"html\" ] del df [ \"image\" ] del df [ \"URL\" ] del df [ \"hash\" ] del df [ \"source\" ] df [ \"label\" ] = df [ \"note\" ] . apply ( lambda x : 0 if x <= 0 else 1 ) df . head ( 5 ) .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } authors note resume texte titre label 0 [Danny Bradbury, Marco Santori, Adam Draper, M... -10.0 Black Market Reloaded, a black market site tha... Black Market Reloaded, a black market site tha... Black Market Reloaded back online after source... 0 1 [Emily Spaven, Stan Higgins, Emilyspaven] 1.0 The UK Home Office believes the government sho... The UK Home Office believes the government sho... Home Office: UK Should Create a Crime-Fighting... 1 2 [Pete Rizzo, Alex Batlin, Yessi Bello Perez, P... -10.0 Though lofty in its ideals, lead developer Dan... A new social messaging app is aiming to disrup... Gems Bitcoin App Lets Users Earn Money From So... 0 3 [Nermin Hajdarbegovic, Stan Higgins, Pete Rizz... 3.0 US satellite service provider DISH Network has... US satellite service provider DISH Network has... DISH Becomes World's Largest Company to Accept... 1 4 [Stan Higgins, Bailey Reutzel, Garrett Keirns,... -10.0 An unidentified 28-year-old man was robbed of ... An unidentified 28-year-old man was robbed of ... Bitcoin Stolen at Gunpoint in New York City Ro... 0 Basic statistics on the dataset let's explore the dataset and extract some numbers : * the number of article liked/disliked df [ \"label\" ] . value_counts () 0 879 1 324 Name: label, dtype: int64 Create the full content column df [ 'full_content' ] = df . titre + ' ' + df . resume #exclude the full texte of the article for the moment df . head ( 1 ) .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } authors note resume texte titre label full_content 0 [Danny Bradbury, Marco Santori, Adam Draper, M... -10.0 Black Market Reloaded, a black market site tha... Black Market Reloaded, a black market site tha... Black Market Reloaded back online after source... 0 Black Market Reloaded back online after source... from sklearn.model_selection import train_test_split training , testing = train_test_split ( df , # The dataset we want to split train_size = 0.75 , # The proportional size of our training set stratify = df . label , # The labels are used for stratification random_state = 400 # Use the same random state for reproducibility ) training . head ( 5 ) .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } authors note resume texte titre label full_content 748 [Jon Brodkin] -10.0 Amazon, Reddit, Mozilla, and other Internet co... Amazon, Reddit, Mozilla, and other Internet co... Amazon and Reddit try to save net neutrality r... 0 Amazon and Reddit try to save net neutrality r... 1183 [Jon Brodkin] -10.0 (The Time Warner involved in this transaction ... A group of mostly Democratic senators led by A... Democrats urge Trump administration to block A... 0 Democrats urge Trump administration to block A... 769 [Joseph Brogan] -10.0 On Twitter, bad news comes at all hours, with ... On Twitter, bad news comes at all hours, with ... Some of the best art on Twitter comes from the... 0 Some of the best art on Twitter comes from the... 57 [Michael Del Castillo, Pete Rizzo, Trond Vidar... -10.0 Publicly traded online travel service Webjet i... Publicly traded online travel service Webjet i... Webjet Ethereum Pilot Targets Hotel Industry's... 0 Webjet Ethereum Pilot Targets Hotel Industry's... 892 [Andrew Cunningham] 10.0 What has changed on the 2017 MacBook, then?\\nI... Andrew Cunningham\\n\\nAndrew Cunningham\\n\\nAndr... Mini-review: The 2017 MacBook could actually b... 1 Mini-review: The 2017 MacBook could actually b... from sklearn.feature_extraction.text import TfidfVectorizer , CountVectorizer from sklearn.svm import LinearSVC , SVC from sklearn.pipeline import Pipeline from sklearn.model_selection import cross_val_predict from utils.plotting import pipeline_performance steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , LinearSVC ()) ) pipeline = Pipeline ( steps ) predicted_labels = cross_val_predict ( pipeline , training . full_content , training . label ) pipeline_performance ( training . label , predicted_labels ) pipeline = pipeline . fit ( training . titre , training . label ) Accuracy = 80.6% Confusion matrix, without normalization [[624 35] [140 103]] import re from utils.plotting import print_top_features from sklearn.model_selection import GridSearchCV def mask_integers ( s ): return re . sub ( r '\\d+' , 'INTMASK' , s ) steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , LinearSVC ()) ) pipeline = Pipeline ( steps ) gs_params = { #'vectorizer__use_idf': (True, False), 'vectorizer__lowercase' : [ True , False ], 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__C' : np . linspace ( 5 , 20 , 25 ) } gs = GridSearchCV ( pipeline , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline1 = gs . best_estimator_ predicted_labels = pipeline1 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline1 , n_features = 10 ) aaa = gs . predict ( testing . full_content ) == testing . label aaa = aaa [ testing . label == 1 ] testing [ \"titre\" ] . iloc [ ~ aaa . values ] #pipeline1.predict([\"windows xbox bitcoin\"]) from sklearn.externals import joblib joblib . dump ( pipeline1 , 'classifier.pkl' ) gs . predict ([ 'Google' ]) array([1], dtype=int64) steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , SVC ()) ) pipeline = Pipeline ( steps ) gs_params = { #'vectorizer__use_idf': (True, False), 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__C' : np . linspace ( 5 , 20 , 25 ) } gs = GridSearchCV ( pipeline , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline1 = gs . best_estimator_ predicted_labels = pipeline1 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline1 , n_features = 10 ) {'classifier__C': 5.0, 'vectorizer__ngram_range': (1, 1), 'vectorizer__preprocessor': <function mask_integers at 0x00000237491B67B8>, 'vectorizer__stop_words': 'english'} 0.711180124224 Accuracy = 71.2% Confusion matrix, without normalization [[153 0] [ 62 0]] --------------------------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-9-3e0781e307fb> in <module>() 25 pipeline_performance(testing.label, predicted_labels) 26 ---> 27 print_top_features(pipeline1, n_features=10) C:\\Users\\Guillaume\\Documents\\Code\\recommandation\\utils\\plotting.py in print_top_features(pipeline, vectorizer_name, classifier_name, n_features) 81 def print_top_features(pipeline, vectorizer_name='vectorizer', classifier_name='classifier', n_features=7): 82 vocabulary = np.array(pipeline.named_steps[vectorizer_name].get_feature_names()) ---> 83 coefs = pipeline.named_steps[classifier_name].coef_[0] 84 top_feature_idx = np.argsort(coefs) 85 top_features = vocabulary[top_feature_idx] C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py in coef_(self) 483 def coef_(self): 484 if self.kernel != 'linear': --> 485 raise ValueError('coef_ is only available when using a ' 486 'linear kernel') 487 ValueError: coef_ is only available when using a linear kernel from sklearn.naive_bayes import BernoulliNB steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , BernoulliNB ()) ) pipeline2 = Pipeline ( steps ) gs_params = { 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__alpha' : np . linspace ( 0 , 1 , 5 ), 'classifier__fit_prior' : [ True , False ] } gs = GridSearchCV ( pipeline2 , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline2 = gs . best_estimator_ predicted_labels = pipeline2 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline2 , n_features = 10 ) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - {'classifier__alpha': 0.25, 'classifier__fit_prior': True, 'vectorizer__ngram_range': (1, 1), 'vectorizer__preprocessor': <function mask_integers at 0x00000237491B67B8>, 'vectorizer__stop_words': 'english'} 0.805900621118 Accuracy = 78.1% Confusion matrix, without normalization [[140 13] [ 34 28]] Top like features: ['use' 'just' 'year' 'price' 'time' 'Bitcoin' 'bitcoin' 'new' 'The' 'INTMASK'] --- Top dislike features: ['ABBA' 'cable' 'cab' 'byte' 'publication' 'bye' 'publications' 'publicity' 'buyer' 'publicizing'] from sklearn.naive_bayes import MultinomialNB steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , MultinomialNB ()) ) pipeline3 = Pipeline ( steps ) gs_params = { 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__alpha' : np . linspace ( 0 , 1 , 5 ), 'classifier__fit_prior' : [ True , False ] } gs = GridSearchCV ( pipeline3 , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline3 = gs . best_estimator_ predicted_labels = pipeline3 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline3 , n_features = 10 ) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - {'classifier__alpha': 0.5, 'classifier__fit_prior': False, 'vectorizer__ngram_range': (1, 1), 'vectorizer__preprocessor': <function mask_integers at 0x00000237491B67B8>, 'vectorizer__stop_words': 'english'} 0.80900621118 Accuracy = 79.1% Confusion matrix, without normalization [[141 12] [ 33 29]] Top like features: ['time' 'Google' 'Pro' 'Apple' 'new' 'The' 'Bitcoin' 'price' 'bitcoin' 'INTMASK'] --- Top dislike features: ['ABBA' 'categories' 'catching' 'catalyst' 'catalog' 'casually' 'casts' 'cast' 'cashier' 'ran']","tags":"Machine Learning","url":"redoules.github.io/machine-learning/Source code for the recommandation engine for articles.html","loc":"redoules.github.io/machine-learning/Source code for the recommandation engine for articles.html"}]}
\ No newline at end of file
+{"pages":[{"title":"About Guillaume Redoulès","text":"I am a data scientist and a mechanical engineer working on numerical methods for stress computations in the field of rocket propulsion. Prior to that, I've got a MSc in Computational Fluid Dynamics and aerodynamics from Imperial College London. Email: guillaume.redoules@gadz.org Linkedin: Guillaume Redoulès Curriculum Vitae Experience Thermomecanical method and tools engineer , Ariane Group , 2015 - Present In charge of tools and methods related to thermomecanical computations. Focal point for machine learning. Education MSc Advanced Computational Methods for Aeronautics, Flow Management and Fluid-Structure Interaction , Imperial College London, London. 2013 Dissertation: \"Estimator design for fluid flows\" Fields: Aeronautics, aerodynamics, computational fluid dynamics, numerical methods Arts et Métiers Paristech , France, 2011 Generalist engineering degree Fields: Mechanics, electrical engineering, casting, machining, project management, finance, IT, etc.","tags":"pages","url":"redoules.github.io/pages/about.html","loc":"redoules.github.io/pages/about.html"},{"title":"Displaying the content of a TABLE","text":"Displaying the content of a TABLE in SQL In this very simple example we will see how to display the content of a table. Since we are working in the notebook, we will load the sql extension in order to manipulate the database. The database mydatabase.db is a SQLite database already created before the example. #load the extension % load_ext sql #connect to the database % sql sqlite : /// mydatabase . db 'Connected: @mydatabase.db' In order to extract all the values from a table, we will use the following command : SELECT * FROM TABLE In our example, we want to display the data contained in the table named tutyfrutty % sql SELECT * FROM tutyfrutty * sqlite:///mydatabase.db Done. index fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 7 Cranberry red 308","tags":"SQL","url":"redoules.github.io/sql/display_table.html","loc":"redoules.github.io/sql/display_table.html"},{"title":"Opening a file with python","text":"Opening a file with python This short article show you how to open a file using python. We will use the with keyword in order to avoid having to close the file. There is no need to import anything in order to open a file. All the function related to file manipulation are part of the python standard library In order to open a file, we will use the function open. This function takes two arguments : the path of the file the mode you want to open the file The mode can be : 'r' : read 'w' : write 'a' : append (writes at the end of the file) 'b' : binary mode 'x' : exclusive creation 't' : text mode (by default) Note that if the file does not exit it will be created if you use the following options \"w\", \"a\", \"x\". If you try to open a non existing file in read mode 'r', a FileNotFoundError will be returned. It is possible to combine multiple options together. For instance, you can open a file in binary mode for writing using the 'wb' option. Python distinguishes between binary and text I/O. Files opened in binary mode return contents as bytes objects without any decoding. In text mode , the contents of the file are returned as str, the bytes having been first decoded using a platform-dependent encoding or using the specified encoding if given. Writing to a file Let's first open (create) a text file a write a string to it. filepath = \". \\\\ myfile.txt\" with open ( filepath , 'w' ) as f : f . write ( \"Hello world !\" ) Reading a file we can now see how to read the content of a file. To do so, we will use the 'r' option with open ( filepath , \"r\" ) as f : content = f . read () print ( content ) Hello world ! A word on the with keyword In python the with keyword is used when working with unmanaged resources (like file streams). The python documentation tells us that : The with statement clarifies code that previously would use try...finally blocks to ensure that clean-up code is executed. In this section, I'll discuss the statement as it will commonly be used. In the next section, I'll examine the implementation details and show how to write objects for use with this statement. The with statement is a control-flow structure whose basic structure is: with expression [ as variable ]: with - block The expression is evaluated, and it should result in an object that supports the context management protocol (that is, has enter () and exit () methods).","tags":"Python","url":"redoules.github.io/python/Opening_file.html","loc":"redoules.github.io/python/Opening_file.html"},{"title":"Opening a SQLite database with python","text":"Opening a SQLite database with python This short article show you how to connect to a SQLite database using python. We will use the with keyword in order to avoid having to close the database. In order to connect to the database, we will have to import sqlite3 import sqlite3 from sqlite3 import Error In python the with keyword is used when working with unmanaged resources (like file streams). The python documentation tells us that : The with statement clarifies code that previously would use try...finally blocks to ensure that clean-up code is executed. In this section, I'll discuss the statement as it will commonly be used. In the next section, I'll examine the implementation details and show how to write objects for use with this statement. The with statement is a control-flow structure whose basic structure is: with expression [ as variable ]: with - block The expression is evaluated, and it should result in an object that supports the context management protocol (that is, has enter () and exit () methods). db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : print ( \"Connected to the database\" ) #your code here except Error as e : print ( e ) Connected to the database","tags":"Python","url":"redoules.github.io/python/Opening_SQLite_database.html","loc":"redoules.github.io/python/Opening_SQLite_database.html"},{"title":"Reading data from a sql database with pandas","text":"Reading data from a sql database with pandas When manipulating you data using pandas, it is sometimes useful to pull data from a database. In this tutorial, we will see how to querry a dataframe from a sqlite table. Note than it would also work with any other sql database a long as you change the connxion to the one that suits your needs. First let's import pandas and sqlite3 import pandas as pd import sqlite3 from sqlite3 import Error We want to store the table tutyfrutty in our dataframe. To do so, we will query all the elements present in the tutyfrutty TABLE with the command : SELECT * FROM tutyfrutty db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : df = pd . read_sql ( \"SELECT * FROM tutyfrutty\" , conn ) del df [ \"index\" ] #juste delete the index column that was stored in the table except Error as e : print ( e ) df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 6 Cranberry red 308 7 Cranberry red 308","tags":"Python","url":"redoules.github.io/python/Reading_data_from_a_sql_database_with_pandas.html","loc":"redoules.github.io/python/Reading_data_from_a_sql_database_with_pandas.html"},{"title":"Writing data to a sql database with pandas","text":"Writing data to a sql database with pandas When manipulating you data using pandas, it is sometimes useful to store a dataframe. Pandas provides multiple ways to export dataframes. The most common consist in exporting to a csv, a pickle, to hdf or to excel. However, exporting to a sql database can prove very useful. Indeed, having a well structured database is a great for storing all the data related to your analysis in one place. In this tutorial, we will see how to store a dataframe in a new table of a sqlite dataframe. Note than it would also work with any other sql database a long as you change the connxion to the one that suits your needs. First let's import pandas and sqlite3 import pandas as pd import sqlite3 from sqlite3 import Error # Example dataframe raw_data = { 'fruit' : [ 'Banana' , 'Orange' , 'Apple' , 'lemon' , \"lime\" , \"plum\" ], 'color' : [ 'yellow' , 'orange' , 'red' , 'yellow' , \"green\" , \"purple\" ], 'kcal' : [ 89 , 47 , 52 , 15 , 30 , 28 ] } df = pd . DataFrame ( raw_data , columns = [ 'fruit' , 'color' , 'kcal' ]) df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 Now that the DataFrame has been created, let's push it to the sqlite database called mydatabase.db in a new table called tutyfrutty db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : df . to_sql ( \"tutyfrutty\" , conn ) except Error as e : print ( e ) except ValueError : print ( \"The TABLE tutyfrutty already exists, read below to understand how to handle this case\" ) Note that if the table tutyfrutty was already existing, the to_sql function will return a ValueError. This is where, the if_exists option comes into play. Let's look at the docstring of this function : \"\"\" if_exists : {'fail', 'replace', 'append'}, default 'fail' - fail: If table exists, do nothing. - replace: If table exists, drop it, recreate it, and insert data. - append: If table exists, insert data. Create if does not exist. \"\"\" Let's say, I want to update my dataframe with some new rows df . loc [ len ( df ) + 1 ] = [ 'Cranberry' , 'red' , 308 ] df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 7 Cranberry red 308 8 Cranberry red 308 I can now replace the table with the new values using the \"replace\" option db_file = \". \\\\ mydatabase.db\" try : with sqlite3 . connect ( db_file ) as conn : df . to_sql ( \"tutyfrutty\" , conn , if_exists = \"replace\" ) except Error as e : print ( e )","tags":"Python","url":"redoules.github.io/python/Writing_data_to_a_sql_database_with_pandas.html","loc":"redoules.github.io/python/Writing_data_to_a_sql_database_with_pandas.html"},{"title":"Creating a sqlite database","text":"Creating a sqlite database When you want to start with using databases SQlite is a great tool. It provides an easy onramp to learn and prototype you database with a SQL compatible database. First, let's import the libraries we need import sqlite3 from sqlite3 import Error SQlite doesn't need a database server, however, you have to start by creating an empty database file import os def check_for_db_file (): if os . path . exists ( \"mydatabase.db\" ): print ( \"the database is ready\" ) else : print ( \"no database found\" ) check_for_db_file () no database found Let's then create a function that will connect to a database, print the verison of sqlite and then close the connexion to the database. def create_database ( db_file ): \"\"\" create a database connection to a SQLite database \"\"\" try : with sqlite3 . connect ( db_file ) as conn : print ( \"database created with sqlite3 version {0}\" . format ( sqlite3 . version )) except Error as e : print ( e ) create_database ( \".\\mydatabase.db\" ) database created with sqlite3 version 2.6.0 check_for_db_file () the database is ready You're all set. From now on, you can open the database and write sql querries into it.","tags":"Python","url":"redoules.github.io/python/Creating_a_sqlite_database.html","loc":"redoules.github.io/python/Creating_a_sqlite_database.html"},{"title":"Setting up the notebook for plotting with matplotlib","text":"Importing Matplotlib First we need to import pyplot, a collection of command style functions that make matplotlib work like MATLAB. Let's, as well, use the magic command %matplotlib inline in order to display the figures in the notebook import matplotlib.pyplot as plt % matplotlib inline # this doubles image size, but we'll do it manually below # %config InlineBackend.figure_format = 'retina' The following parameters are recommended for matplotlib, they will make matplotlib output a better quality image # %load snippets/matplot_setup.py plt . rcParams [ 'savefig.dpi' ] = 300 plt . rcParams [ 'figure.dpi' ] = 163 plt . rcParams [ 'figure.autolayout' ] = False plt . rcParams [ 'figure.figsize' ] = 20 , 12 plt . rcParams [ 'axes.labelsize' ] = 18 plt . rcParams [ 'axes.titlesize' ] = 20 plt . rcParams [ 'font.size' ] = 16 plt . rcParams [ 'lines.linewidth' ] = 2.0 plt . rcParams [ 'lines.markersize' ] = 8 plt . rcParams [ 'legend.fontsize' ] = 14 plt . rcParams [ 'text.usetex' ] = False # True activates latex output in fonts! plt . rcParams [ 'font.family' ] = \"serif\" plt . rcParams [ 'font.serif' ] = \"cm\" plt . rcParams [ 'text.latex.preamble' ] = \" \\\\ usepackage{subdepth}, \\\\ usepackage{type1cm}\" You can change the second line in order to fit your display. 163 dpi corresponds to a Dell Ultra HD 4k P2715Q. You can check your screen's dpi count at http://dpi.lv/","tags":"Python","url":"redoules.github.io/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html","loc":"redoules.github.io/python/Setting_up_the_notebook_for_plotting_with_matplotlib.html"},{"title":"Why using a blockchain is a bad idea for your business","text":"What having a blockchain implies? storage costs : everyone maintaining the ledger needs to store every transaction bandwith costs : everyone has to broadcast every transaction computational costs : every node has to validate the blockchain control : the creator does not control the blockchain, everyone collectively controls it developpement costs : developping on a blockchain is way harder than on a traditionnal database What to ask a business when they tell you that they are using a blockchain? When a business is telling you about their innovative technology leveraging the power of the blockchain this should immedialty spake some questions : What is the consensus algorithm? who is responsible for validating the consensus rules? what is the nature of the participation ? is it open to access? is it open to innovation? is it a public ledger? is it transparent? does it improves acountability? is it cross borders? how is it validated?","tags":"Cryptocurrencies","url":"redoules.github.io/cryptocurrencies/blockchain_bad.html","loc":"redoules.github.io/cryptocurrencies/blockchain_bad.html"},{"title":"Synology NFS share","text":"Setting up a NFS share login to your DSM admin account, open the \"Control Panel\" and go to \"File Services\" Make sure NFS is enabled Back in the control panel, go to \"Shared Folder\" Select the folder you want to share and clic \"Edit\" Go to the \"NFS Permissions tab and clic \"Create\", add the ip of the device you want to mount the mapped drive on. Make sure you copy the \"Mount path\"","tags":"Linux","url":"redoules.github.io/linux/share_nfs_share.html","loc":"redoules.github.io/linux/share_nfs_share.html"},{"title":"Mount a NFS share using fstab","text":"Mount nfs using fstab The fstab file, generally located at /etc/fstab lists the differents partitions and where to load them on the filesystem. You can edit this file as root by using the following command sudo nano /etc/fstab in the following example, we want to mount a NFS v3 share from : server : 192.168.1.2 mountpoint (on the server) : /volumeUSB2/usbshare * mountlocation (on the client) : /mnt we specify 192 .168.1.2:/volumeUSB2/usbshare /mnt nfs nfsvers = 3 ,users 0 0 the client will then automatically mount the share ont /mnt at startup. Related you can reload the fstab file using this method : https://redoules.github.io/linux/Reloading_fstab.html You can create a NFS share on a Synology using the method : https://redoules.github.io/linux/share_nfs_share.html","tags":"Linux","url":"redoules.github.io/linux/mount_nfs_share_fstab.html","loc":"redoules.github.io/linux/mount_nfs_share_fstab.html"},{"title":"Installing bitcoind on raspberry pi","text":"Installing bitcoind on linux Running a full bitcoin node helps the bitcoin network to accept, validate and relay transactions. If you want to volunteer some spare computing and bandwidth resources to run a full node and allow Bitcoin to continue to grow you can grab an inexpensive and power efficient raspberry pi and turn it into a full node. There are plenty of tutorials on the Internet explaining how to install a bitcoin full node; this tutorial won't go over setting up a raspberry pi and using ssh. In order to store the full blockchain we will mount a network drive and tell bitcoind to use this mapped drive as the data directory. Download the bitcoin client Go to https://bitcoin.org/en/download Copy the URL for the ARM 32 bit version and download it onto your raspberry pi. wget https://bitcoin.org/bin/bitcoin-core-0.15.1/bitcoin-0.15.1-arm-linux-gnueabihf.tar.gz Locate the downloaded file and extract it using the arguement xzf tar xzf bitcoin-0.15.1-arm-linux-gnueabihf.tar.gz a new directory bitcoin-0.15.1 will be created, it contrains the files we need to install the software Install the bitcoin client We will install the content by copying the binaries located in the bin folder into /usr/local/bin by using the install command. You must use sudo because it will write data to a system directory sudo install -m 0755 -o root -g root -t /usr/local/bin bitcoin-0.15.1/bin/* Launch the bitcoin core client by running bitcoind -daemon Configuration of the node Start your node at boot Starting you node automatically at boot time is a good idea because it doesn't require a manual action from the user. The simplest way to achive this is to create a cronjob. Run the following command crontab -e Select the text editor of your choice, then add the following line at the end of the file @reboot bitcoind -daemon Save the file and exit; the updated crontab file will be installed for you. Full Node If you can afford to download and store all the blockchain, you can run a full node. At the time of writing, the blockchain is 150Go ( https://blockchain.info/fr/charts/blocks-size ). Tree ways to store this are : use a microSD with 256Go or more add a thumbdrive or an external drive to your raspberry pi * mount a network drive from a NAS If you have purchased a big SD card then you can leave the default location for the blockchain data (~/.bitcoin/). Otherwise, you will have to change the datadir location to where your drive is mounted (in my case I have mounted it to /mnt) In order to configure your bitcoin client, edit/create the file bitcoin.conf located in ~/.bitcoin/ nano ~/.bitcoin/bitcoin.conf copy the following text # From redoules.github.io # This config should be placed in following path: # ~/.bitcoin/bitcoin.conf # [core] # Specify a non-default location to store blockchain and other data. datadir=/mnt # Set database cache size in megabytes; machines sync faster with a larger cache. Recommend setting as high as possible based upon mach$ dbcache=100 # Keep at most <n> unconnectable transactions in memory. maxorphantx=10 # Keep the transaction memory pool below <n> megabytess. maxmempool=50 # [network] # Maintain at most N connections to peers. maxconnections=40 # Tries to keep outbound traffic under the given target (in MiB per 24h), 0 = no limit. maxuploadtarget=5000 Check https://jlopp.github.io/bitcoin-core-config-generator it is a handy site to edit the bitcoin.conf file Pruning node If you don't want to store the entire blockchain you can run a pruning node which reduces storage requirements by enabling pruning (deleting) of old blocks. Let's say you want to allocated at most 5Go to the blockchain, then specify prune=5000 into your bitcoin.conf file. Edit/create the file bitcoin.conf located in ~/.bitcoin/ nano ~/.bitcoin/bitcoin.conf copy the following text # From redoules.github.io # This config should be placed in following path: # ~/.bitcoin/bitcoin.conf # [core] # Set database cache size in megabytes; machines sync faster with a larger cache. Recommend setting as high as possible based upon mach$ dbcache=100 # Keep at most <n> unconnectable transactions in memory. maxorphantx=10 # Keep the transaction memory pool below <n> megabytess. maxmempool=50 # Reduce storage requirements by only storing most recent N MiB of block. This mode is incompatible with -txindex and -rescan. WARNING: Reverting this setting requires re-downloading the entire blockchain. (default: 0 = disable pruning blocks, 1 = allow manual pruning via RPC, greater than 550 = automatically prune blocks to stay under target size in MiB). prune=5000 # [network] # Maintain at most N connections to peers. maxconnections=40 # Tries to keep outbound traffic under the given target (in MiB per 24h), 0 = no limit. maxuploadtarget=5000 Checking if your node is public one of the best way to help the bitcoin network is to allow your node to be visible and to propagate block to other nodes. The bitcoin protocole uses port 8333, other clients should be able to share information with your client. Run ifconfig and check if you have an ipv6 adresse (look for adr inet6:) IPV6 Get the global ipv6 adresse of your raspberry pi Link encap:Ethernet HWaddr xx:xx:xx:xx:xx:xx inet adr:192.168.1.x Bcast:192.168.1.255 Masque:255.255.255.0 adr inet6: xxxx::xxxx:xxxx:xxxx:xxxx/64 Scope:Lien adr inet6: xxxx:xxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx/64 Scope:Global UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:42681744 errors:0 dropped:0 overruns:0 frame:0 TX packets:38447218 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 lg file transmission:1000 RX bytes:3044414780 (2.8 GiB) TX bytes:2599878680 (2.4 GiB) it is located between adr inet4 and Scope:Global adr inet6: xxxx:xxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx/64 Scope:Global Copy this adresse and past it into the search field on https://bitnodes.earn.com/ If your node is visible, it will appear on the website IPV4 If you don't have an ipv6 adresse, you will have to open port 8333 on your router and redirect it to the internal IP of your raspberry pi. It is not detailed here because the configuration depends on your router.","tags":"Cryptocurrencies","url":"redoules.github.io/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html","loc":"redoules.github.io/cryptocurrencies/Installing_bitcoind_on_raspberry_pi.html"},{"title":"Reloading .bashrc","text":"Reload .bashrc The .bashrc file, located at ~/.bashrc allows a user to personalize its bash shell. If you edit this file, the changes won't be loaded without login out and back in. However, you can use the following command to do it source ~/.bashrc","tags":"Linux","url":"redoules.github.io/linux/Reloading_.bashrc.html","loc":"redoules.github.io/linux/Reloading_.bashrc.html"},{"title":"Reloading fstab","text":"Reload fstab The fstab file, generally located at /etc/fstab lists the differents partitions and where to load them on the filesystem. If you edit this file, the changes won't be automounted. You either have to reboot your system of use the following command as root mount -a","tags":"Linux","url":"redoules.github.io/linux/Reloading_fstab.html","loc":"redoules.github.io/linux/Reloading_fstab.html"},{"title":"Updating all python package with anaconda","text":"Updating anaconda packages All packages managed by conda can be updated with the following command : conda update --all Updating other packages with pip For the other packages, the pip package manager can be used. Unfortunately pip hasn't the same update all fonctionnality. import pip from subprocess import call for dist in pip . get_installed_distributions (): print ( \"updating {0}\" . format ( dist )) call ( \"pip install --upgrade \" + dist . project_name , shell = True )","tags":"Python","url":"redoules.github.io/python/updating_all_python_package_with_anaconda.html","loc":"redoules.github.io/python/updating_all_python_package_with_anaconda.html"},{"title":"Saving a matplotlib figure with a high resolution","text":"creating a matplotlib figure #Importing matplotlib % matplotlib inline import matplotlib.pyplot as plt import numpy as np Drawing a figure # Fixing random state for reproducibility np . random . seed ( 19680801 ) mu , sigma = 100 , 15 x = mu + sigma * np . random . randn ( 10000 ) # the histogram of the data n , bins , patches = plt . hist ( x , 50 , normed = 1 , facecolor = 'g' , alpha = 0.75 ) plt . xlabel ( 'Smarts' ) plt . ylabel ( 'Probability' ) plt . title ( 'Histogram of IQ' ) plt . text ( 60 , . 025 , r '$\\mu=100,\\ \\sigma=15$' ) plt . axis ([ 40 , 160 , 0 , 0.03 ]) plt . grid ( True ) plt . show () Saving the figure normally, one would use the following code plt . savefig ( 'filename.png' ) <matplotlib.figure.Figure at 0x2e45e92f400> The figure in then exported to the file \"filename.png\" with a standard resolution. In adittion, you can specify the dpi arg to some scalar value, for example: plt . savefig ( 'filename_hi_dpi.png' , dpi = 300 ) <matplotlib.figure.Figure at 0x2e462164898>","tags":"Python","url":"redoules.github.io/python/Saving_a_matplotlib_figure_with_a_high_resolution.html","loc":"redoules.github.io/python/Saving_a_matplotlib_figure_with_a_high_resolution.html"},{"title":"Iterating over a DataFrame","text":"Create a sample dataframe # Import modules import pandas as pd # Example dataframe raw_data = { 'fruit' : [ 'Banana' , 'Orange' , 'Apple' , 'lemon' , \"lime\" , \"plum\" ], 'color' : [ 'yellow' , 'orange' , 'red' , 'yellow' , \"green\" , \"purple\" ], 'kcal' : [ 89 , 47 , 52 , 15 , 30 , 28 ] } df = pd . DataFrame ( raw_data , columns = [ 'fruit' , 'color' , 'kcal' ]) df .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } fruit color kcal 0 Banana yellow 89 1 Orange orange 47 2 Apple red 52 3 lemon yellow 15 4 lime green 30 5 plum purple 28 Using the iterrows method Pandas DataFrames can return a generator with the iterrrows method. It can then be used to loop over the rows of the DataFrame for index , row in df . iterrows (): print ( \"At line {0} there is a {1} which is {2} and contains {3} kcal\" . format ( index , row [ \"fruit\" ], row [ \"color\" ], row [ \"kcal\" ])) At line 0 there is a Banana which is yellow and contains 89 kcal At line 1 there is a Orange which is orange and contains 47 kcal At line 2 there is a Apple which is red and contains 52 kcal At line 3 there is a lemon which is yellow and contains 15 kcal At line 4 there is a lime which is green and contains 30 kcal At line 5 there is a plum which is purple and contains 28 kcal","tags":"Python","url":"redoules.github.io/python/Iterating_over_a_dataframe.html","loc":"redoules.github.io/python/Iterating_over_a_dataframe.html"},{"title":"Article Recommander","text":"import pandas as pd import numpy as np % matplotlib inline Loading data and preprocessing we first learn the pickled article database. We will be cleaning it and separating the interesting articles from the uninteresting ones. df = pd . read_pickle ( './article.pkl' ) del df [ \"html\" ] del df [ \"image\" ] del df [ \"URL\" ] del df [ \"hash\" ] del df [ \"source\" ] df [ \"label\" ] = df [ \"note\" ] . apply ( lambda x : 0 if x <= 0 else 1 ) df . head ( 5 ) .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } authors note resume texte titre label 0 [Danny Bradbury, Marco Santori, Adam Draper, M... -10.0 Black Market Reloaded, a black market site tha... Black Market Reloaded, a black market site tha... Black Market Reloaded back online after source... 0 1 [Emily Spaven, Stan Higgins, Emilyspaven] 1.0 The UK Home Office believes the government sho... The UK Home Office believes the government sho... Home Office: UK Should Create a Crime-Fighting... 1 2 [Pete Rizzo, Alex Batlin, Yessi Bello Perez, P... -10.0 Though lofty in its ideals, lead developer Dan... A new social messaging app is aiming to disrup... Gems Bitcoin App Lets Users Earn Money From So... 0 3 [Nermin Hajdarbegovic, Stan Higgins, Pete Rizz... 3.0 US satellite service provider DISH Network has... US satellite service provider DISH Network has... DISH Becomes World's Largest Company to Accept... 1 4 [Stan Higgins, Bailey Reutzel, Garrett Keirns,... -10.0 An unidentified 28-year-old man was robbed of ... An unidentified 28-year-old man was robbed of ... Bitcoin Stolen at Gunpoint in New York City Ro... 0 Basic statistics on the dataset let's explore the dataset and extract some numbers : * the number of article liked/disliked df [ \"label\" ] . value_counts () 0 879 1 324 Name: label, dtype: int64 Create the full content column df [ 'full_content' ] = df . titre + ' ' + df . resume #exclude the full texte of the article for the moment df . head ( 1 ) .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } authors note resume texte titre label full_content 0 [Danny Bradbury, Marco Santori, Adam Draper, M... -10.0 Black Market Reloaded, a black market site tha... Black Market Reloaded, a black market site tha... Black Market Reloaded back online after source... 0 Black Market Reloaded back online after source... from sklearn.model_selection import train_test_split training , testing = train_test_split ( df , # The dataset we want to split train_size = 0.75 , # The proportional size of our training set stratify = df . label , # The labels are used for stratification random_state = 400 # Use the same random state for reproducibility ) training . head ( 5 ) .dataframe thead tr:only-child th { text-align: right; } .dataframe thead th { text-align: left; } .dataframe tbody tr th { vertical-align: top; } authors note resume texte titre label full_content 748 [Jon Brodkin] -10.0 Amazon, Reddit, Mozilla, and other Internet co... Amazon, Reddit, Mozilla, and other Internet co... Amazon and Reddit try to save net neutrality r... 0 Amazon and Reddit try to save net neutrality r... 1183 [Jon Brodkin] -10.0 (The Time Warner involved in this transaction ... A group of mostly Democratic senators led by A... Democrats urge Trump administration to block A... 0 Democrats urge Trump administration to block A... 769 [Joseph Brogan] -10.0 On Twitter, bad news comes at all hours, with ... On Twitter, bad news comes at all hours, with ... Some of the best art on Twitter comes from the... 0 Some of the best art on Twitter comes from the... 57 [Michael Del Castillo, Pete Rizzo, Trond Vidar... -10.0 Publicly traded online travel service Webjet i... Publicly traded online travel service Webjet i... Webjet Ethereum Pilot Targets Hotel Industry's... 0 Webjet Ethereum Pilot Targets Hotel Industry's... 892 [Andrew Cunningham] 10.0 What has changed on the 2017 MacBook, then?\\nI... Andrew Cunningham\\n\\nAndrew Cunningham\\n\\nAndr... Mini-review: The 2017 MacBook could actually b... 1 Mini-review: The 2017 MacBook could actually b... from sklearn.feature_extraction.text import TfidfVectorizer , CountVectorizer from sklearn.svm import LinearSVC , SVC from sklearn.pipeline import Pipeline from sklearn.model_selection import cross_val_predict from utils.plotting import pipeline_performance steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , LinearSVC ()) ) pipeline = Pipeline ( steps ) predicted_labels = cross_val_predict ( pipeline , training . full_content , training . label ) pipeline_performance ( training . label , predicted_labels ) pipeline = pipeline . fit ( training . titre , training . label ) Accuracy = 80.6% Confusion matrix, without normalization [[624 35] [140 103]] import re from utils.plotting import print_top_features from sklearn.model_selection import GridSearchCV def mask_integers ( s ): return re . sub ( r '\\d+' , 'INTMASK' , s ) steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , LinearSVC ()) ) pipeline = Pipeline ( steps ) gs_params = { #'vectorizer__use_idf': (True, False), 'vectorizer__lowercase' : [ True , False ], 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__C' : np . linspace ( 5 , 20 , 25 ) } gs = GridSearchCV ( pipeline , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline1 = gs . best_estimator_ predicted_labels = pipeline1 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline1 , n_features = 10 ) aaa = gs . predict ( testing . full_content ) == testing . label aaa = aaa [ testing . label == 1 ] testing [ \"titre\" ] . iloc [ ~ aaa . values ] #pipeline1.predict([\"windows xbox bitcoin\"]) from sklearn.externals import joblib joblib . dump ( pipeline1 , 'classifier.pkl' ) gs . predict ([ 'Google' ]) array([1], dtype=int64) steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , SVC ()) ) pipeline = Pipeline ( steps ) gs_params = { #'vectorizer__use_idf': (True, False), 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__C' : np . linspace ( 5 , 20 , 25 ) } gs = GridSearchCV ( pipeline , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline1 = gs . best_estimator_ predicted_labels = pipeline1 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline1 , n_features = 10 ) {'classifier__C': 5.0, 'vectorizer__ngram_range': (1, 1), 'vectorizer__preprocessor': <function mask_integers at 0x00000237491B67B8>, 'vectorizer__stop_words': 'english'} 0.711180124224 Accuracy = 71.2% Confusion matrix, without normalization [[153 0] [ 62 0]] --------------------------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-9-3e0781e307fb> in <module>() 25 pipeline_performance(testing.label, predicted_labels) 26 ---> 27 print_top_features(pipeline1, n_features=10) C:\\Users\\Guillaume\\Documents\\Code\\recommandation\\utils\\plotting.py in print_top_features(pipeline, vectorizer_name, classifier_name, n_features) 81 def print_top_features(pipeline, vectorizer_name='vectorizer', classifier_name='classifier', n_features=7): 82 vocabulary = np.array(pipeline.named_steps[vectorizer_name].get_feature_names()) ---> 83 coefs = pipeline.named_steps[classifier_name].coef_[0] 84 top_feature_idx = np.argsort(coefs) 85 top_features = vocabulary[top_feature_idx] C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\svm\\base.py in coef_(self) 483 def coef_(self): 484 if self.kernel != 'linear': --> 485 raise ValueError('coef_ is only available when using a ' 486 'linear kernel') 487 ValueError: coef_ is only available when using a linear kernel from sklearn.naive_bayes import BernoulliNB steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , BernoulliNB ()) ) pipeline2 = Pipeline ( steps ) gs_params = { 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__alpha' : np . linspace ( 0 , 1 , 5 ), 'classifier__fit_prior' : [ True , False ] } gs = GridSearchCV ( pipeline2 , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline2 = gs . best_estimator_ predicted_labels = pipeline2 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline2 , n_features = 10 ) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:820: RuntimeWarning: divide by zero encountered in log neg_prob = np.log(1 - np.exp(self.feature_log_prob_)) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:823: RuntimeWarning: invalid value encountered in add jll += self.class_log_prior_ + neg_prob.sum(axis=1) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:801: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - {'classifier__alpha': 0.25, 'classifier__fit_prior': True, 'vectorizer__ngram_range': (1, 1), 'vectorizer__preprocessor': <function mask_integers at 0x00000237491B67B8>, 'vectorizer__stop_words': 'english'} 0.805900621118 Accuracy = 78.1% Confusion matrix, without normalization [[140 13] [ 34 28]] Top like features: ['use' 'just' 'year' 'price' 'time' 'Bitcoin' 'bitcoin' 'new' 'The' 'INTMASK'] --- Top dislike features: ['ABBA' 'cable' 'cab' 'byte' 'publication' 'bye' 'publications' 'publicity' 'buyer' 'publicizing'] from sklearn.naive_bayes import MultinomialNB steps = ( ( 'vectorizer' , TfidfVectorizer ()), ( 'classifier' , MultinomialNB ()) ) pipeline3 = Pipeline ( steps ) gs_params = { 'vectorizer__stop_words' : [ 'english' , None ], 'vectorizer__ngram_range' : [( 1 , 1 ), ( 1 , 2 ), ( 2 , 2 )], 'vectorizer__preprocessor' : [ mask_integers , None ], 'classifier__alpha' : np . linspace ( 0 , 1 , 5 ), 'classifier__fit_prior' : [ True , False ] } gs = GridSearchCV ( pipeline3 , gs_params , n_jobs = 1 ) gs . fit ( training . full_content , training . label ) print ( gs . best_params_ ) print ( gs . best_score_ ) pipeline3 = gs . best_estimator_ predicted_labels = pipeline3 . predict ( testing . full_content ) pipeline_performance ( testing . label , predicted_labels ) print_top_features ( pipeline3 , n_features = 10 ) C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - C:\\Users\\Guillaume\\Anaconda3\\lib\\site-packages\\sklearn\\naive_bayes.py:699: RuntimeWarning: divide by zero encountered in log self.feature_log_prob_ = (np.log(smoothed_fc) - {'classifier__alpha': 0.5, 'classifier__fit_prior': False, 'vectorizer__ngram_range': (1, 1), 'vectorizer__preprocessor': <function mask_integers at 0x00000237491B67B8>, 'vectorizer__stop_words': 'english'} 0.80900621118 Accuracy = 79.1% Confusion matrix, without normalization [[141 12] [ 33 29]] Top like features: ['time' 'Google' 'Pro' 'Apple' 'new' 'The' 'Bitcoin' 'price' 'bitcoin' 'INTMASK'] --- Top dislike features: ['ABBA' 'categories' 'catching' 'catalyst' 'catalog' 'casually' 'casts' 'cast' 'cashier' 'ran']","tags":"Machine Learning","url":"redoules.github.io/machine-learning/Source code for the recommandation engine for articles.html","loc":"redoules.github.io/machine-learning/Source code for the recommandation engine for articles.html"}]}
\ No newline at end of file