From 7a7df0b11bf1f40442dfca829fdc31181da96f41 Mon Sep 17 00:00:00 2001 From: Arthur DANJOU Date: Thu, 7 Mar 2024 15:44:04 +0100 Subject: [PATCH] Remove .Rhistory & .RData --- Analyse Multidimensionnelle/DM ACP/.RData | Bin 19783 -> 0 bytes Analyse Multidimensionnelle/DM ACP/.Rhistory | 512 ------------------- Analyse Multidimensionnelle/TP1/.RData | Bin 39809 -> 0 bytes Analyse Multidimensionnelle/TP1/.Rhistory | 155 ------ 4 files changed, 667 deletions(-) delete mode 100644 Analyse Multidimensionnelle/DM ACP/.RData delete mode 100644 Analyse Multidimensionnelle/DM ACP/.Rhistory delete mode 100644 Analyse Multidimensionnelle/TP1/.RData delete mode 100644 Analyse Multidimensionnelle/TP1/.Rhistory diff --git a/Analyse Multidimensionnelle/DM ACP/.RData b/Analyse Multidimensionnelle/DM ACP/.RData deleted file mode 100644 index 3ced7c0cc3ffb42acde204d2630292511153ee1f..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 19783 zcmV)5K*_%!iwFP!000001MR&BR20v$IK1Q_Ihx5igQy@P{4#r22qkrP>djA zL{v~gF@uVN5kx>_$clnQ$sjp{WQpO0oU^C5duqC?tE;NJs(Yq3 z>26hF-pWiM5NHUrG}Htdy2-x;I#ZL?a?1$>TI$IR6@iw(G*xKj>geK5AW*YSX4s&~ zZu)}nq!eMyp1u%HmFWxNBTQe&{ORHxMBv)gbsZ7L?u$<6uOZ?S(-$H&U5zawGY$U{ zk)H-rLKO9<3uYlhW8>g7nXpd&ri$67ZxIOJXMwTEZ?bf@adDD&v$eG$!BZ0n1j(?; z|08DR_jd1F?|Zd=Sc21U7a`4${!W3VwVzFv=kkYN%vi@g!W8w%zAScpk&b(Cb`S^AYXmGwVdBB zNpW-M7LKM5X~wQ2tKEZCih7iuCTyt*-(7k1&}Ty?TLaD&^^P9Qu_qlnG*^|mteL;- z{tmO}JIbpim0MnZk=wEINi(xa4;xJiUyI_(P7+M>n$(Rs*2Ov*U-C|5&q>tlEC3ahhe&R>fx_dUCu zH(xDNzU1z*GW)Yz+S|3*wm*5uxNsvsufXo1D;{I!;VfQD8|y~eH@_7?_AQgh{9J7o zI%lriYwANTo{ICf=d?s{=6~26kQ$qRKYpq2E9FkEQ2qnqV=Qs@`U=Bu$9uV#bSDm7 zCstQCIy1)|6E*CmH`u-N;+eMu6$LMseCFmxZ$qn=UHsPX9x?6wZgw=yYR{Y#hAb}? zTMdU!6wul8dMD-+9tB$Qr=CAt=lYIn6W!A*dpQ_wJ!t1-H6^#+xq`-7XUw(B313S^ zI5!cryX~U(lfW(g@o%WUWpL4To|jvFA}MkCGi|Zmhvde{opGlCk>-o4bLA}3kuq~pH*{9 zn)*&n&zEus?iXsB&3Uyg{(eda;~Xi^SnYVf9@gL+B4O73#Qhcz(iJSdPe~im?xYEM zY?A9jIQZ92$*@1PY zFHT1dSATuBYVHc#PpdXAJZ@b^)OfSDbFTbi3D@lg1|osP2(2?lkrLpQ6~D@~^zMf6 zlV>HfEE|_TkfhU2ikiR9IWcL3vvPK~BeE>M2{t>jlTj!jvyMa(2_la4=?b<@AKr(8{S~7x2!jc~v4QY+Ug~^2E7)EajP-Gz_oo z?xi}~(S39Rbi=%NY}HDUR~PqQ-mNRXI+4Lun8##Ee`@(ob&J!1N{7|@I`Y_<6@`;c zc_Ma7Dfttvs1jxA{go9@K11TlKg0wH=ICv=i%sC}{P#Tj-nku`5@ zj`SsleztXC0*}o<${xfoBSd#dfSV*nQYk` z)Ka&TYvk!tL!-eXMzZOf^4-M$?do4>&~s+`UsUf=)_)O%%~r>)N0eF&3UxXns*Y-SuyPTpM7OwFZlV&6z5RXpQu!y>qjh=n zZr<3J$jUVT^4h&ik3Uo$PImM%*SONK`eLhkny^>4Jfr2ahb*o41#Yl04{_~J^YvC$ z2Od~1C26^3NZrl;VP3s}w&QgV-6fjsvVoD-suoL8OZ^W=Eo(N9y}su)VIGl2tMAV$ zB$JrS+V*JdO0}lhMq^o9;>sf%wSwNgRMvj(xCQwnasS$AYIDL)uc0ev&3fe`oOT#< ziW&&hoTl=ijZ)56&Xz+@UD=Y8e@q13TozunL`bE7RsHRS7t|FJ7@2~L#eH`g6bgPx zR?SJx4J=vcVzJL^e%9+dPe!e7cxB)EqPYFhJtxDm`tW%-KeWj=JUXrMRb%CoM;V%1 zG&^xYV5YV77fD4cFD+qu{C(Bp!qWHI3@Z|Eu6c1+B7SVW(UAkHUD52sP+=wa@g02&+A=&B zo~~!6R!U6097dZ_8qT%fVbhZ}JP#M`%W&sAlu)}>e^P`#@A37cjGkYsJhw&6E-wD@!g(zFVJ=ztXUFQ~Y zwxxVH=hN2Gwr)}WsB3Q8eZ7$MdGnfy9dpaq6dm5f*_IP_^@{0({+s(5{mPfD%xx@S zeD66ZWGDJEDCo08<-q3B!v)vYvT8UDEhq9iGox7a~V8q=Il#Pa%3tvXWVg(HEf;U=~-(%c54@BccO!pxs!{Rxu+!& zUlt+@1J{ZMt`(mJ&h6lAiR;$TesiI%SBL|aS8SvP2{EZuCe zGFEK)$wpd&PL)lslBE%;5KFq8>W_#(YBA4Q?xV=O?mIDW+*crTPjMMZd^(8mW>&V` z;W~!!K6xH}wG%R^hM z3W`Rk;yA9RxG}@`#@GdyWFNx!DkJd*t`r!pP+viOU=u-wU#y1rkDsFp2<5ew5=3paeNZ-3ZFuxkA`9FnKx-&z` zk<-hUQnf;U;}zkj7ib}8j%kzFw+u)-5=XU8_Ab2M##TwFn}8P*2RTnZc@6jMcO9x? zz6-bX3-z8|w1;DZqsKYdSAb=$bN_g@70lsM^VyZ3I@=_1wsf*}oBYEuMeM23n*1@Z zv30ifbV!Tvw8cv3wH%z%9^pAgowP@ewr=jWSS9r9JRBWto$+DaU`e#bYucRli0I<% zu+Ia3mRZYY@?p!gXEt~V-CF0g2si9u24jzdmd?(2&2)OU&O`@1qwituK)$oi#gUw0 zFmSMTakRuMVV=am-ouvY;Qj;J<0o|2Sh~5p;OH`~@o;c-wDfQy!y4NX?ePrbbl2Q} zJgY@?nzV;-Lg-EQB`t!CF2fpIm&s?{@Sbh5bxwPP!=|0Faq+Mu+Tb-!_Xfv;-jrx% z>44`L*15R-)F^|Nv%8Cfv+a+X+$W#>(W;@NgQtTn5g#i(TcYP=vslH(UXGqWs5VI% za)w!FH_>5|60W;{>YcNt2i`m8)kI6@w1+MZZg|T~Cc7ucb8@uEsH~gH$2k2U?IbO2 z-N@B>5M7Amwv8-39PtdTzKe$|ChOBC=#&LZ(g9Qa$v8n_NeWB;%UE*eH0NaLI5~Z} z5^ddVo!xESre;TsJIPa=p{u)ti?h73>ffFsl?mi2l751eGGOUn%z~;7QYZN07!U>I z7@~R591*KA>kduah6ucrwamSF1mP{Ld!T;!Hp0QoqaG{14q>bPwyw6+17WM8c9v`i zK^TvmhJdl92;BwiGPCrA1Ad5H(hxdLd8 zYl)3`vK*R}-=4HMG!JTelUcduqflcTc6UH}36#?^mv=a>hvF8o5UZO6D2gffx)^;H z3VBzAZ6I_)o^6!b?QU22)bvWLbnF2ni$5A)`Fah!`Pxu^z-uWaTsS1PL&pFjBD^9i z_K89St&iSaiyFA+vve%&o(|lm`uK`bb^@+3jdr)MI1iT#cJC~oUjcyH3S7`5mwPB8Iax0WNsI>X@1QK7-4C=I}qMpP!Z8?^Mf7QJkSj$b$D9;ljoH zclvXN#+#vTe$`yFRNG8dJ5y_rWa_t?=!c^ERkVMrCrEmCw&wbsS|iT|zgJE(l=e)` zOIB;MRqd~H$4q@PTlf5;oMvmYzdi9$Sdzk$|7@0=X&YF(5Y4w@Q_wuQ5rAAcUdkHM2VNj*XM zvG+ko`pxJ(qm6?TZX0o9a2(Tc7}<=7E>>iQn;axdM=vr>C$sjBmJVd8-JoS4Z|rXA zjx!qXbVh!yv)$xaOfD)MY^P=XpV7kXL8hOS8y6zpHa*!icO>%@4bc`?D0KEj+aF@( zHCd<2emv*uL3Aan6nd9Q$(fYbnQd`;S`z=)Yaupd*_)*gI=VuUC0pF$`iGcjEg7x+*khlVb|F~8_jP4_Z*JoHTG)fiW@LMiy6&ZmrBrVrqc(o6q z^)|gzNM8pl__yV#|3iEl9eMOjWi0s?Kd|71Q67G}2d{tzh7w+rA zd}CxE5$>y+?Nh>iO_;BX>~q6>X=I-l?z6&uU6{`a_nH0To5Ou;vwVDH9~tgL!+c+4 z-yi1dBm1~89~$nX!+mJD50C6K!+m48wS!wLv#lQ78o{j^+{(bL7_!xcSy8wZM7GK> z>x*nXVb%xP8o{j}%o@V2H?nnuTR*sUG~1fOtR38{!>uy1b%lYxa zT|(tqNhkEJJ6%wi;}3OJDfSv-@1dpi~>jI|=upeXacd+og7s<0vRTk;(`j=tSnjo@hifj!Sv<9zDn}3r(Ek^bKCz z${H-F_kujFi4YO*JJ8uD=&G;#9zTB^;Uol> z?QV`kL~<|KQLzp}W3uIfS|xR8+q=}{pkOLO$3NF*>FF+nz-4_wccCA2-LUjgjeZ6t z%N8)b*!2w|e2Z&ev}iumR|E~+PBKBLpF3Web8i$bhzMJ<7zaQ`&r5!_Ynf0#f9Wdj z1Q)2;Xq&9a7KSjaDSFfInuIJKs}YMO#$Wly8902N|}5 zbH17@K{3M)4Lv^Woua-lE3I#>-Q9RFgAPZfusj)A1L`1l2w|K=rL#Tp_R>YO^B1|$a+j#QZpmO1X zt7Qi);YIF3{XwPmkoCg$5_dx_BDSt~|HY&3P;wxH7__V$@)d28Ct^jx zcZm-BE${D;Ws|+wO;-WFOVjGk=O2Wj<7n!E9e^l```7O(A4eAN4ovgv(S!@@D$EzN zuSJA7pRH|;Xn~p)G4}kc`Jk$T*Y^F~A%uP;{1v^pFmxJp8kXFug>X9~AD=JL5cY1l zoapHaguBtbe;wN$XmI%OCS0Hk!ajX*;tM$eF~#zprGZ>9zWtiD`S~}H`QbHh0E-7y zdveIz&3y&Yhgd@#i=X~x2`&9&3C%JwYsrk8vNJuNGfcF*n@NS@Ig zle5c|cY3m#$wl>){ISoHKWeQ#_&*cvU(3kHUIt4g=8=wp3Sh@T z6|nO8*zwT$SUn4{a#1Xm#^5Ef`o;B0$4Z5;yd>#7Dk6p*H(iW9HxJ9pVC{-w?Z{#G z#W6gE2s18339DCXRyi-0;&_W=aEe&{N*JEX$)s`t3{DusPa3BOUO(2p7?u~497w&e zLyAmS9l;LLOG%E>$?(c?n^)*}gD}Q3;WzwvO~CHWc(c8bjvZUGtRZ z;N+w&5SYpP$qf-~9c^|zr;BjCd@h(C`T*fq%sutxm;qtdg^1j&s;VGVBJQeEy{! zMGI2OkIXEmU%s*3Z*d$VoKrB>VMOY_h1Qv6`iL+Fuk;>E)nv_2xPK+{hqz!!Tkm)P zB75OfSww^>A{ctV-uPQCBDT8lj!|$QvLM_q$7db;Pfd~NN`sA)%H@N#lN3Y5R=S3j zFt0@BP4-%HAPtel#!Cw8mk7p}@>nV|4Ze@e$3oN9976b}9#6f12zyLbG>9lo4~HHi ziouC6Pd9|{Og-)zj!0wm%*V?4Fg#@ANc4~;;X|g=TWi>S6~{QO^71af6)k>PiNAVnTJGA1*{zb?0Fn- z84`Yo9F`XxAkj$%lLJvwzYxJ&Q(yv!K1dMIDOxd=q4i;9r?CxMUORQEP^;K$U-?Z{!{A~|nnITuDJT(3!E<1K~JQxUr_ zi|HFtzLdS)L1Ty{hL5-z-E-%KyUF@Z7N_gXamDFMZdc_2362ZHdjU3X^4m!5;PouP z?#sE6Qkt3lfMs zinXVL%pTI2<10wIkH};Flf(M0ggw6mgIk2<7h-UVjx+IBV8QUm=O^Q7yvY1I55p7J z+w-vccrkcfugmx0&ynht!P?`Ug*Pi!4=47VNb=0_6~pMX2)obejo~@79vH&lr7$?* zK5SmW`XhwVoddhifX}Dmq~~Zz&;LSK39MdeY+UhqNO?7>Tty9Q4}+7$##5GMCS7GP ze1!1%lswKPy7FS8Sh*;cid@9`hg9u^c zB*BOA*TtFj$OmD1EQ*AOTrz%(MAtc({#3Lk>956@o|VG*QGp%P-&py4Z2pnK^xpz( zUg5yn6U6i+3#RWyF?jhAk{%Po`GAhpt{8^D%-d=GO~#uKpRX{zDwj{1FC?*fU4n!c zbYXD#Jgcymq<6VU`2Ipy$!^m8D22^GlDNLPLF%6jrq9K(dgU;FRK(^NNlYKhV)#p7 z_vOj@kVIE;On*vX^N}c?Mw0MTav{-GZr3!u$oPq0n#SuFyz%*jgS4JQ#MDW6$Yc28 z>tKm3(-RE&Iq6w+qOm~zFzYx=BQVV}e&mVNf=%L6! z8gD_2U*t*rh6rQpKXDR1h&VP*Quw?nPwIyZmX~=&qL(mfya==BEhUWq_^|O1#OR`E zNTRDW)=zBxi->1F{Ce}H+-$s|51WtKuzG|sI!a@FDva^J5XQfvB5?r)hzX)d;%_oO zG*~@S7@vq28*kIut%^ut{4IcuAI{f{Fn&;)xeoY+u40({DB$ZGELFzntBC1cag6_k zBQXB3oUzV9q%nAGorNgP(y!te9Bzl;dPi|qUfKKi_myVlaeYTSt6U7vWA$Kq9=D4S zWh}+l{nE4YIG(atia#fb*N@Aq_snv}KlBU7lM736J5F*IoYbuQ%HA{cINd}D_(h_t zBsQ<%av+9}7naIm{DH4`aJy8#Z05R1!hgnmNiN6nlf>q6^1P8e10U=)$X}nA!P>+5 z9;dI&wwZRVD0W{8>yJ3g@9QIvr}xZq1#G>gh>-H)7=E%?JNW#M(-|+9!}>**e2lL6 z_~R*#H%MQ7zDXTi~*&1#bP|5Lxo zbevT$PFGxRWoA9EI4hN&b^lLvoR$C6MeY~QFZlXuwjNUUo(YHdUwIZBx&JeE{n>GE z=Z2}B@SjgL{k&!V*Bkvm9TE8Be*Ir;&;PvZ|3BEF|K%*r-#k|EkM`u!Svy*~;qGF* zklDu4-BR9;Xz4WjSPc`=WxxE)_c$mUE|d+If9i(I?0pg{{M_JTbg=wx#xt%ZFzB&u z%riCwefe%*&LPsci*!uD8k}GIsJI})YXCqK^ z(&S#zU==z4EoI(gF5wEac79V%41)`rTP3NNe|H|*@;xuQf8I5;EnLH8aL*eU8Ff-k z>C%IyB+4B1EI6~9IR#tCD8ffs& z3L8vq`OW4iJ>k!rqxAn5vieUrNH}A(S>3EH9c^vYt*7?6h}g)}A>u+LPrAsoWo?hh znNDAb)J__Qyrnl0i5ExL^EX7#nlb*l=O0-x)Ta|N=5ux0x|h9_M)+F2l;h;-5!z|q zx`fdChq7kU@gh9ACO6)0TmR3-{1@zJys_yp{SpxAjZGRB#WWVhj9K6u)-EnY8#nVw zC2j6TYLnDqXz`{gyp(hneg9IEb9aO|yz3iWr2TLZO+7KeaXCa0t&V^8h7eBkhYtTk zw^&(;Q)iL1Npa^@k6 zHqMbEBz#BMS!;RTE%=Hs9z12*{%$uym3}dhp4$e-qfH}E^zWOUJM0J+`O!Oapd-lVve5kg+4?-G>4;$nvbA){1W}$1+1eS{(bn6FY$IUTFY0QF&NP2eSNI`LdTSRaS3#v2 zIprBSl^MAuGjggka!Y6Amd(g5pOI6;Wr)@hl%&hXvtjf&n60Acs_?&9%U~lW} zY5C*aGoPz1aY~_C+S>}c*a>=0YF$fbYy7T|wTF|3qouopr>)>00H;}-dWw3q827G* zHj(AThXSXe?Q!bHqiliDd_2$9mtPc`lP-PlC{Kp=29~96%9YT$ld;Hsz8SO?zufiy z^g-x!5G7u+?ZxF3I&3Cx>mPT4Qs#~QCeM~Y6UXu7z17w8_UdWl;YTWv zYLa!ji2Xj43`^%3`u9Td9vQmzQGHNr%SP?dD2aXuN*>rBYKLZoKVHVEWr^l)Uo<2T zcN9h6Hc*+I=EC&>nqSvi)~o*o>aspq9+&zA=6iblm#(Kq3!)Ws72|o(ik0$8KGloS z6p10<)oER5vVKp8$e0VH%CzY5dM5vTJ zejE+6x#G3BWEEMz#n&>|DYBTL<&BA%NI)N2YZEFKZ?Xq!txm7G^7;_^Xk?jW2)87B zy3qb1S85sh=3v?3#DiIoOmo({`1mWh+Z?}KZ9yo6d{wKl$Yw`lWsd}w>*_5yJ$rmw67QrO>?JP;&hDq{%Op?1XNj^18l9ez? zuEiwzJSNFXm?Vc|lH7wyaxW&y;MCSa1xg-P-POp-e> zNp{90nG2I-xml8|jY+aUCdrF2N$#5^$uBWUmc=Cb8YamrFiEDxBzZR`$#R$^6K6?s z1t!TyF-g|NBw2ozBzI$y?0`wK7beMKm?W!WlKc*nw5`JYRtorpLIg*YGIojn*yQsmPKYt&KhVj zl*{TG3Wc_wr-$uRw?NbJS zM8N~DwD&6oE#T(rDlaC2E_QTnsg>`Mi`HE{Ev|R6ALL zI|QOLV*RJLu81VpfYR3uzcfqUVepB(L$LQIYL zg36%95Y-_+-+k8w*lN{$KCS}k`ckW zM^5G}mWWVLfbX1l%aH}f+&(6odyoayk6uOH*^h{*)9k-;q8br1JhCX`E*m10^XRND zT?ry)$^E%*-6>>Ifvt;{_^G_{$?WfrXl`@A^rv;{)Qp`1|sv=5DxV+q1!zA2tUpJ&Fi_FAwT5I z%Pixo2<;l{+95&*w6r!KtYk=!SG9T-d1db%%X1bgK4NtlMdT@Kc|tRBzsdFiLM8 zGBk5WmMf_58+0_GNb^fkfuva|q!q zwefX2PpFnZ?QS4_3ySBD2%0LWqqU)INi|t(5L!*zI)TBYWKP06<|AEOlbLRrzucCm zjL^)1%m|@Y=!wWGzZPN!qh7+_*6`kev$>Wk$7NWO>7KtldV%XvGJB6zv{Dc~!tPLg zd-tLJ$#jDgN=ddR$@4iQqz*P6Nai`6)K#ZZ0>c_FsQuqjBh=dKi8SW&(4H|b(dgrC zgexGTt9nox;TY`R>+vxS(ys;W%28{EZ{3vv&*a(>y1YGQbLMfNePbmlvMTNfeKpH` zg~ldmLX(;obml=3+unl?W;@UxqR;*p*_WWvPiMg#HIV4*7H^V!6|HJf2o(P!4`044 zD>-%)OH%Icl zEa64Jtofv~*>^n*i@&7*!i~V^uQhG)H!7eltx#K9G!yDx8-1~m+zyQg{gqAJ!_e-K zBg~tURAIzV%5N~G11%He=qW!hi#EBQZ@21t3-y}!M7iQ$p)VL|Y?mpC!1&JXua3@% zfj$w2dlg$a(epd%ZpA;aMK66$V(lv(fvO9Q-~8%7PcC)mU(J2+1{wtq?cc7QheqwS z6Np=0n9NZx$#Ug_I+P`KNgo*9k7ivqe;1Lq7A*~2Fqp9X9NK)jkZ*gv1^Ru=cZ+(( zS9EOfc4ey!Kl<(>`?IJUmhb?HxW6ieA8ny-SsXs>hxR>Ky=N%F6KedkGBorm(cbcb zOVvZCpdo12CHbTr=*_*E$lW`J4h#A;HeN>&8o3YW)qS5s!`z5cKhHR563Y`0ym|ud z%PRlmqGO46X70@lDe{Geb6NZ%doE0lTUE|uNiMW2c=+nMx*e14Sf2Ty(T{esmK3aw((J}X1 zkApi_!540}co%{TS~a?B<%;W%(R{gZszH}{D19KQ>%R2?+BF6eOWRxF+rYhd`;;C- z-G|lIJL1;D=VQd9yZBJ(c>GdJPbU-|8xqU7P*aOG32=9rC4`_od$Mg6Eq9@V+{eye zG0lSfMSKV9=kTMgD{qJSpVfqFUAqlO+DxIg#GlJ9ZXr~k*x*iog@7jZK9@YqvkjSZ z^+IK|lNz)edHY8TB|__ZwvC$&1JK6qPowTra_GEO?GCRU+OavTEbQnY`aH2Zz4A#b zlqaPtwoOE#EjBwAHQdiczXb&%AwKcFy_at54k~&Z(umKuOup|M%X{ z|A!lA1yHv+uj2rXD0IeMdOfIf5K5O3L#tPcLrU=W)t&j0(4iu-fb(25G*t_4xcgZe zisboDZ+KAxEg!n0vN-lb+4>#ba?7%y{@jrH+rz@p(D6P(eC;==T^i8UYaRq^7+m7# zz$>WpG_8{Ta0uv)cxbzqxI0S zvFZL!?;yyxUvtH0yb0PPm#mLm(+r))#n)`opF#dHqmEba%Ahg3?nV8>Qg}eOjzx3* z9Ox)Xsu+3o1lkPmZ4@_bhJb!%K0bPVC_7eFdFOyG)ZP>DtXrB6i9QqRJZJ)R%sIN> zqHhrLu2g0w6x)M)zcyjum>9&weq&~f^n)wHF1(AiPO!%iro42^LcAbM)Ejz;oq;q?EJm5YBYk_TYI|xWW7}c4KP- zB;Tr{YqhWdVmiw<-f#7ArHm$_$}<1@H3K4Cpe@`JWtsqQ6O*_DCHKKLrWjdg zUv7x;k~^?OClek-TP3^e#XvE&6%CYR!d($btsFihxZxA}MpWR`WWORdR7~pH%Uy*B z-vrNvlqFH7oIO#nXLmkd$jBL3A6auI((yU?6!Ape$>xL8IyJVh8n3_uRRsqb$-9ub zb@d^~Yo6ecf0oM4RarBrV0-rk?t{>zW9j|jqa!?hY<+RNUm!>ulAmUh6U~hCTJnf9!nAgh+&qqs^e_mn(8HORE ztnLfp^s=6eL-Xq)ur&Z;&pN{0Z}dXf&FVoX=(C2+0ad7f6x`6S7zqtbCUe;u(t*Y( z@IY>MK2TlAdD41p4frsBHjLgY1NTe=n?02m!*gkB#%?xAFrb}Bm2GGX4Gwni^`f`K zlPuYG@sc-S7v*v8#lmXvDrD&6BRF9 z)O=W`QxyQ`R%-|G>v2HbJ!H+EcUq9lQY*&?pP?|!^BsLQGu&CvwK@UW1>wh>!pcsc zg*##*g62z;Aar@%)*Y1_A;G)fL}76`#9F_a8(zH_t_4un3uVoPi;35yiDS|5>8#aj z+lK9s#CV=MD25X<#%;8ZG^N5TdR3Leo_!EgZStkpY8SZX+I1GL@rSFUo6Mxbset%E z-0US?Je(9&b6(cu4A-s%$ai1#0W;e#eLiX1pz%#Sa;eq>Zfou-v!9;;H#Lo%mDqzJ zKj3UzCGQACGLDYqUZ{m9sbZ;p9-*Kz9=!#*yA*=tJ)LPM2BDDCXt?%+HY9#uB`Y~< z1U{U=AEkK%PVbCdyUNZNG6OczJ$6xrySj59B&)21#~k7=5nb(IFcI^3>4u~5{`Qu` z+96%woBT90j_ne-4zl-^ogzTIoGj5~C>1iS*l!F6G(eES{AStR*WtzIHvx9LF2T9i z+Vz14q`^KRDE20U7+gM`tNg7e7%ILPJWZ2+559Ch+~?SDmniWWI*toUv1ox2>SSzgQol>GTECE-eCSji(v! zw<*JmafMWS=`pZ2l)P-ww+)<4!%vAkE`^u-&TZfOffqu*S6>rrYXdFXBwDB5y|7$4 zDPcv%dw3kU!tYFSFuY1$mYE&h1KUuV0}4agusNV^qipOxxOI)+{st8a`zjP}xTWod zC_kHH)LQwl;eJMNWUL=$4Mtgm{lB&bqkKq=@*y$Ghr}o!5~F-bjPfBd%7?@#YcR?h z?4P~{lSbGc*Y+QZP(!%lciuQPE{o9ct&(2xVi!XDVG+Yw*)<3YAM(WcG#kS5-mCoJ zr84O?;Un5M< z5XWv|X@o9Q?XtZ`AVU3y{^I*>hA^a{!d3Tm5klSglJoYtc?cu7xard2K7>b$_xnKw zBEp(Tlpib4N9bq=!lIaqVSLN?fzgwp2(RdHfoGQ;j66G@uUO)Uuvt4veq!8+aGRVr zTk)U-;dr@}c@fkgyzIx<(+5i+3z=y3qmtGlylzaj_7|NHMq#e8HpwQ0UO+~5S#c!7 zk@&UGd0q~}eXfJA3f+Qmy*f%gynzM!dkJ%ky|obP!JL*B_pLGP#aqFcn&p<9*ng#MX%&>YN8ALAhc4L#A) zZF^orZ7E^tKK?-U8)tL;WR@_@K>fM02g|c|fEzbgnKe-d?Wz*m zo1TQCgYmukC!OP9aNB7@c9AOjRZHq(wWfbEwLkyaI7St8d^vS;s=F!F_6N2nj^2aT z(dBjtU<-vQjdmL=I$*@UjzgZf0ft24smiP};ECaYcokOqDj=pz$j-ifFTfw!)FtP_a=6+vUqPF*MYMLl8%AUK7x~|TRB81B znp0E`xqes;od=$5c`ufO7JWaScyJ&C?cZ6`tWvTX&Dp=Ng+@CY604t8&WQVxMI04nY=L#u_XfmVL2%+K(0b0*y zCdqK*qJ>oZ1T{Z8qvbc51-A)ph5NN%D0no^cxRmnG-aITeRZi2zLYk-xwz#x z)UB4v)>6?%J65W7c3NA&;KgO{u3S8fezvd+JbL6LTF<|K#X*&Ps4>4R9wMNKMvY9U zl-5_n&}iCW)s2PFt;avKZayb^dUU6e?v?Xsz&4Kb{(-YlCZ4{8jg}wk#ce}$inAui z@r-b(ttT2G*FrqayegR`QIF30yeNEbzR=AvVuEI*ZJ(IK>4_FK-;Pqz`iwSk2yw2S z2tY@6jMB#(F+_*LcR1fre~+fKa9``_pMZN^ueZG8RY4n^vJO)na74Rbr+f-H5dxJ` z8|7N<@1otznGuS&jV8^ogFLE?iO@|Y$UQ%_6dkb0jM8TRh!8eH*>L%GsGY~XZh!Jx zs4J;aOI+WKb{8`&FXMlTcG`5V-hA*hH0T|F;c?g&EghvEIs3*2?NZ6^>MBcxn$5~W zT+)xwuP^;eGHz%amazwVo*SKB#McL7 zl{S{Qo}Wj@_=%c>GNw@d#E><;f{0dKQ{Gi=VU6Yu*aVfTMMJ46@p22bI@)#hgY2$s zPx$swflrHR6V&;mrQK#b;IkT=4(+SG(4k+HUfL>wj$K>#w6HH0Z7jY|vp7o!?NLgP z_3KnZ2R<(KPd6@wyhhfQ1DZ8xOWW{70v97x)vn(Vs;&jKHw6tols(PP4 za+x-BQ?%hVOWNK=%#-*o^PjNVfTk_26|B1(0K;x^Uo7g*L5W^w<+xceT6|2);a&b2 z7=0iR{VX6UnLwj{aHFvcbfr+aJQJsaUK@9WOZyO7ab%&(mA79YqjoXdmBS-w!4A5G~v7`X#|&hatw>YOUHz-wo$;qG9x z!|Tl5@k%MQHIhy&e_s|`8mSTL3RM)@*`@tLr4yjKt3*L~UbU0M^ZI(t*+ z_CzUKX`*mGU_2h}32c;&cMV7DOF!>bG2w^gfZBC~H-pjS4S6C@^;f|+K^@*NXKT@# zi|bDAEA574qi+^@HGARRV}GI4Ay%}5Z|LMg69&llKJ}TSY8hI2b4$C3?xfjRH;?9# zlJlR<#Gg*_7ld5(5!pfm1;ex*)QjgsUV_q^MP`bS*AO!?xAYDamb8vVc?CgHgJA02 zPy;BOyT+fGNDD*8>h(wk_? zfS{!*UpM-JX4AE1bx|~7^ZW_FQBO4eNT{CvKt7uD)X>D9y$8O?Im zE@CjbjZSX4-v%Y)2x+FI$dHnm+{Pr6F`4;U=ltIFo_D?PTAydHwV&_)W3RpT+W$Py zrq$DOz>S3}`=4Li`B(F*DpCqUuJ__aWfONsSI#&I6(Q4xTxUea82eoo&p&HTvlHZR zs)lA{iLOpPOT4m0hAz)FW6ig2`A(A7pfPGboSp#JwZPpGT3{k<|4@4T_K>qnsGV)m z^0A#Y|5Hz`VK~RR6GtRJXA`h};_$ZyAFhD%(?;2H=-kZ155G_q`aJ@BGH1rklH|G# z_=4+f6r^S=&MHt!l=;ig->pv9yw5(-K%Uk^Eb{m#kD&wi;@-vqV2%3(XUGSh%?{{Y zbVG5Cy;<8uS8Q(hs&3~XssytFY!L62>F2$v7M6S!hd&gEfuD=H<(V-WzfK2My-RZ@ zpZ-d(eNH;({t)ownp_g&YMMt_$?o>-qW4VHu4eUgU48N9vl6*}-iZ&Ey4tm$;Yr<*BC+wotY z&5Mv;QJ>plwH=qXV@5?hOsOQdkbuAnDwbUIuul4moa09#p^3`JoUpwo#ZW|eOYQxB z)br}wCi93uXHobf&D3y^amI1HSvP2KHluhxL}6*7Dfe-6LpB%c_F8vSX5NWIxm<|a z{(4<^o~fCS7F05%xa~rg52cd&3@K&vlBisvplKKiVa4 zhBu#zYHv7xttrhF4=(uQ3gV-e(_|i+82SEj(gYFNqQY;1-!Nnj2?sZKy*L_U{ETFP zjtZlApl9O#0gCxOu?Q4K_QAG5suvw+|{Z=AU;{-9| z6JiNM>J2I3C~R+|?;))a1bajorhwTEo@tDLr;bf=wzQaUBUWFFH2YB)+;g8$-aEC~ z@jc3=A9gVo0~3+yVP{R91ARemhiBhORhpnl-dmJ{sLyh;)Ei$~cPW4l(U%T0oq#)| z2CLdIcBElT1tOYUPR#p3X#UVSbwpMtVe|aWby!rWcBP8*Sk;v!RZT-#_(Q{%tp#mz zOHlSgWo16URM3y`nUbF0B{p^nlk<>c;@Wd-ov1U5^5{acM<^gkSU^jq^R=I9!wAn@)q+N?vd>v7exF@N2j$n8&yw@;yrR=HbXzQ{V%>33hvzZ~p? zZCzIUZ3fC*oT0d|>Sa=7uxD!(N4_n4Occ%+eOD;FhV%ys?z2_v-al{w+#0E_8Mf5B z`#BwbLlZpM_SK5t$l27h`8W)=*N5JQTY!|Jy8JDCjcn7;$zaS*dt}~l+TJ6&Bm*a3fh%})(d|#A>zf~>H@qB+)uHrSpp5)>v1?yTzSuddH zs~og5N<24vUMKLuWA^8kUZL11PvWPh8X?^@@6UwzML@GD55g#H&`vriL3u)F0kcKz z|G94 znTQ6jtTE7b?4ISBpU@>cH{mjMK(WeC(!J^TxdQB^Pw;*0B5s_K*fFGD&pHBp>%8%1 z-_2?Jje)zT0pCgJT*4f5in+m+9}j;F3G;~03+Zg{h>2B{_zm~9iXZvFL7hil?LTq* z+lT254u303Ir>j(izG;?5~QWTg!8=6JQoM6yIk^FVcHh``-NKdGRFC%buLHiI3WjT z#@-y1J|wsntPA zS)<(JLm7I>UA$*5TR&AQ702ts#_ag7v#bw&8$@t1CMPZ=nAEd3*Bc1R2k9I0%*+O`UtRDxh z>v&cA#4f=`8fdnDt29?Q9~V;~4zQoRQEd`^27gCX_m7$AK|iZwSCEjB56W4G^Yuj) z^qKSN=uH(i_-dvbgwlv=U5n0qjZoVSenCs1E7mAkoTqksjOjbSQA{jn zmynaY^)|Hw_DPKA2mhB}Ii=bu*~ndv7;=A>IwlR5i6Wp!I-+8Pt0SD_R+XMl1Ulsk zDqZyHh>l;Eo3v*M=Mz^;a0186h~Zek(i^s9O@&1{qrUi?NSeb=jXk#`D z=^we zADakxG+S<`{YV|75@Svu9+sy9m`fFwM{g#gPb+5hLFaS5HZ?GCPA^^I(g*X zXoNRSi4hVS&05=)*UQl>7fmZnYH3*zh2o?U&x`4J-7xU3)UQ>om;toQQzj)V7^6iyvI!nr1*H&(pS!D5XAY}na>E*gPsEq(M zC(oSp%fSCeevklpv(F+hboc9HJCJQ{t;!!IslXUwyXXrhsp^7TTz9bkbS!=W5Pj+P zZUq_gqFCS?azeJEJ(_Ol!VA`t0BgNQ#kwmKk{#YF{5Mj6 zUjZ;JPt3sUV*fspyose)N$^-umGoj{%d?N_HVVaL#1+mM+U` zZWy9>mP9^n4ewY<;q928J%>Uqo9@vN)7Mb+t29@+*5Db<)KT;O^7-7W zq*swy@&Yjvq=5T!a<4qvVFH6il}r6j zYYI%zuRl}X%XQ)2B|n5VQqMe1MKEDAuBI|)bmLwZ7os9S?{+K07@U6gm3QfgX=t$c zPG-Z@j;v-vtUdDuxmOp}-s0%IyPY3t7Wd2&sR6oMQzld616@gFophnpE>)=v9myRa z?%&Og;BM0Ndk-^&$5((HGXcBV@|w`83lORrHmc~Q>G(?dFaA+EZKkTX8B)dd|DgcY x>;Cd@;RO#uid-O6j@@zps7>!dQ#;`~r`bVvH)RmSdd{`He2P!RR)y!#{{a+ST+;vm diff --git a/Analyse Multidimensionnelle/DM ACP/.Rhistory b/Analyse Multidimensionnelle/DM ACP/.Rhistory deleted file mode 100644 index d152650..0000000 --- a/Analyse Multidimensionnelle/DM ACP/.Rhistory +++ /dev/null @@ -1,512 +0,0 @@ -contribs <- res.notes$ind$contrib -contrib_moy_ind <- mean(contribs) # 100 * 1/42 -contrib_therese <- res.notes$ind$contrib["Thérèse",3] -contrib_moy_ind -contrib_therese -quali_julien <- res.notes$ind$cos2["Julien", 1:2] -quali_julien -sum(quali_julien * 100) -contrib_moy_ind <- mean(res.notes$ind$contrib) -contrib_moy_ind -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,1], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,2:2], decreasing = TRUE) -indiv_contrib_axe_1[1:3] -mal_representes <- rownames(res.notes$ind$contrib)[rowSums(res.notes$ind$contrib[,1:2]) <= contrib_moy_ind] -cat("Les individus mal représentés sont : ", mal_representes) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -summary(res.notes_sup, nb.dec = 2, nbelements = Inf, nbind = Inf) -contrib_moy_var <- mean(res.notes_sup$var$contrib) # 100 * 1/14 -contrib_moy_var -var_contrib_axe_2 <- sort(res.notes_sup$var$contrib[,2:2], decreasing = TRUE) -head(var_contrib_axe_2, 3) -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -indiv_contrib_axe_1[1:3] -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -notes_MAN_prep <- notes_MAN[,-1] -X <- notes_MAN[1:6,]%>%select(c("Probas","Analyse","Anglais","MAN.Stats","Stats.Inférentielles")) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -# View(X) -X <- scale(X,center=TRUE,scale=TRUE) -X -cor_X <- cor(X) -eigen_X <- eigen(cor_X, symmetric = TRUE) -lambda <- eigen_X["values"]$values -vect <- eigen_X["vectors"]$vectors -lambda -inertie_total_1 <- sum(diag(cor_X)) # Inertie est égale à la trace de la matrice de corrélation -inertie_total_1 -inertie_total_2 <- sum(lambda) # Inertie est aussi égale à la somme des valeurs propres -inertie_total_2 -inertie_axes <- (lambda[1] + lambda[2]) / inertie_total_1 # Inertie expliquée par les deux premières composantes principales -inertie_axes -C <- X %*% vect -C[,1:2] -plot( -C[,1],C[,2], -main="Coordonnées des individus par rapport \n aux deux premières composantes principales", -xlab = "Première composante principale", -ylab = "Deuxieme composante principale", -panel.first = grid(), -col = c('blue', 'red', 'green', 'yellow', 'purple', 'orange'), -pch=15 -) -legend(x = 'topleft', legend = rownames(X), col = c('blue', 'red', 'green', 'yellow', 'purple', 'orange'), pch = 15) -nrow(notes_MAN_prep) # Nombre d'individus -ncol(notes_MAN_prep) # Nombre de variables -dim(notes_MAN_prep) # On peut également utiliser 'dim' qui renvoit la dimension -library(FactoMineR) -# help(PCA) -# Ne pas oublier de charger la librairie FactoMineR -# Indication : pour afficher les résultats de l'ACP pour tous les individus, utiliser la -# fonction summary en précisant dedans nbind=Inf et nbelements=Inf -res.notes <- PCA(notes_MAN_prep, scale.unit = TRUE) -summary(res.notes, nbind = Inf, nbelements = Inf, nb.dec = 2) -eigen_values <- res.notes$eig -bplot <- barplot( -eigen_values[, 1], -names.arg = 1:nrow(eigen_values), -main = "Eboulis des valeurs propres", -xlab = "Principal Components", -ylab = "Eigenvalues", -col = "lightblue" -) -lines(x = bplot, eigen_values[, 1], type = "b", col = "red") -abline(h=1, col = "darkgray", lty = 5) -coords_man_stats <- res.notes$var$coord["MAN.Stats",] -coords_man_stats[1:2] -contribs <- res.notes$ind$contrib -contrib_moy_ind <- mean(contribs) # 100 * 1/42 -contrib_therese <- res.notes$ind$contrib["Thérèse",3] -contrib_moy_ind -contrib_therese -quali_julien <- res.notes$ind$cos2["Julien", 1:2] -quali_julien -sum(quali_julien * 100) -contrib_moy_ind <- mean(res.notes$ind$contrib) -contrib_moy_ind -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,1], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 3) -mal_representes <- rownames(res.notes$ind$contrib)[rowSums(res.notes$ind$contrib[,1:2]) <= contrib_moy_ind] -cat("Les individus mal représentés sont : ", mal_representes) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -summary(res.notes_sup, nb.dec = 2, nbelements = Inf, nbind = Inf) -contrib_moy_var <- mean(res.notes_sup$var$contrib) # 100 * 1/14 -contrib_moy_var -var_contrib_axe_2 <- sort(res.notes_sup$var$contrib[,2:2], decreasing = TRUE) -head(var_contrib_axe_2, 3) -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 4) -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 3) -View(X) -colors <- c('blue', 'red', 'green', 'yellow', 'purple', 'orange') -plot( -C[,1],C[,2], -main="Coordonnées des individus par rapport \n aux deux premières composantes principales", -xlab = "Première composante principale", -ylab = "Deuxieme composante principale", -panel.first = grid(), -col = colors, -pch=15 -) -legend(x = 'topleft', legend = rownames(X), col = colors, pch = 15) -View(X) -View(notes_MAN) -View(res.notes) -mal_representes <- rownames(res.notes$ind$contrib)[rowSums(res.notes$ind$contrib[,1:2]) <= 4.334] -cat("Les individus mal représentés sont : ", mal_representes) -mal_representes <- rownames(res.notes$ind$contrib)[rowSums(res.notes$ind$contrib[,1:2]) <= contrib_moy_ind] -cat("Les individus mal représentés sont : ", mal_representes) -mal_representes <- rownames(res.notes$ind$cos2)[rowSums(res.notes$ind$cos2[,1:2]) <= mean(res.notes$ind$cos2[,1:2])] -cat("Les individus mal représentés sont : ", mal_representes) -View(notes_MAN_prep) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot(res.notes_sup, habillage = "Mentions") -knitr::opts_chunk$set(include = FALSE) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -notes_MAN_prep <- notes_MAN[,-1] -X <- notes_MAN[1:6,]%>%select(c("Probas","Analyse","Anglais","MAN.Stats","Stats.Inférentielles")) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -# View(X) -X <- scale(X,center=TRUE,scale=TRUE) -X -cor_X <- cor(X) -eigen_X <- eigen(cor_X, symmetric = TRUE) -lambda <- eigen_X["values"]$values -vect <- eigen_X["vectors"]$vectors -lambda -inertie_total_1 <- sum(diag(cor_X)) # Inertie est égale à la trace de la matrice de corrélation -inertie_total_1 -inertie_total_2 <- sum(lambda) # Inertie est aussi égale à la somme des valeurs propres -inertie_total_2 -inertie_axes <- (lambda[1] + lambda[2]) / inertie_total_1 # Inertie expliquée par les deux premières composantes principales -inertie_axes -C <- X %*% vect -C[,1:2] -colors <- c('blue', 'red', 'green', 'yellow', 'purple', 'orange') -plot( -C[,1],C[,2], -main="Coordonnées des individus par rapport \n aux deux premières composantes principales", -xlab = "Première composante principale", -ylab = "Deuxieme composante principale", -panel.first = grid(), -col = colors, -pch=15 -) -legend(x = 'topleft', legend = rownames(X), col = colors, pch = 15) -nrow(notes_MAN_prep) # Nombre d'individus -ncol(notes_MAN_prep) # Nombre de variables -dim(notes_MAN_prep) # On peut également utiliser 'dim' qui renvoit la dimension -library(FactoMineR) -# help(PCA) -# Ne pas oublier de charger la librairie FactoMineR -# Indication : pour afficher les résultats de l'ACP pour tous les individus, utiliser la -# fonction summary en précisant dedans nbind=Inf et nbelements=Inf -res.notes <- PCA(notes_MAN_prep, scale.unit = TRUE) -summary(res.notes, nbind = Inf, nbelements = Inf, nb.dec = 2) -eigen_values <- res.notes$eig -bplot <- barplot( -eigen_values[, 1], -names.arg = 1:nrow(eigen_values), -main = "Eboulis des valeurs propres", -xlab = "Principal Components", -ylab = "Eigenvalues", -col = "lightblue" -) -lines(x = bplot, eigen_values[, 1], type = "b", col = "red") -abline(h=1, col = "darkgray", lty = 5) -coords_man_stats <- res.notes$var$coord["MAN.Stats",] -coords_man_stats[1:2] -contribs <- res.notes$ind$contrib -contrib_moy_ind <- mean(contribs) # 100 * 1/42 -contrib_therese <- res.notes$ind$contrib["Thérèse",3] -contrib_moy_ind -contrib_therese -quali_julien <- res.notes$ind$cos2["Julien", 1:2] -quali_julien -sum(quali_julien * 100) -contrib_moy_ind <- mean(res.notes$ind$contrib) -contrib_moy_ind -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,1], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 3) -mal_representes <- rownames(res.notes$ind$cos2)[rowSums(res.notes$ind$cos2[,1:2]) <= mean(res.notes$ind$cos2[,1:2])] -cat("Les individus mal représentés sont : ", mal_representes) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.iris, choix = "ind", habillage = "Mentions", label = "none") -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.notes_sup, choix = "ind", habillage = "Mentions", label = "none") -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.notes_sup, choix = "ind", habillage = "Mentions") -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.notes_sup, choix = "ind", habillage = "Mention") -View(notes_MAN) -View(notes_MAN) -knitr::opts_chunk$set(include = FALSE) -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -mal_representes_moy <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= mean(res.notes_sup$var$cos2[,1:2]) -mal_representes_moy -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -notes_MAN_prep <- notes_MAN[,-1] -X <- notes_MAN[1:6,]%>%select(c("Probas","Analyse","Anglais","MAN.Stats","Stats.Inférentielles")) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -# View(X) -X <- scale(X,center=TRUE,scale=TRUE) -X -cor_X <- cor(X) -eigen_X <- eigen(cor_X, symmetric = TRUE) -lambda <- eigen_X["values"]$values -vect <- eigen_X["vectors"]$vectors -lambda -inertie_total_1 <- sum(diag(cor_X)) # Inertie est égale à la trace de la matrice de corrélation -inertie_total_1 -inertie_total_2 <- sum(lambda) # Inertie est aussi égale à la somme des valeurs propres -inertie_total_2 -inertie_axes <- (lambda[1] + lambda[2]) / inertie_total_1 # Inertie expliquée par les deux premières composantes principales -inertie_axes -C <- X %*% vect -C[,1:2] -colors <- c('blue', 'red', 'green', 'yellow', 'purple', 'orange') -plot( -C[,1],C[,2], -main="Coordonnées des individus par rapport \n aux deux premières composantes principales", -xlab = "Première composante principale", -ylab = "Deuxieme composante principale", -panel.first = grid(), -col = colors, -pch=15 -) -legend(x = 'topleft', legend = rownames(X), col = colors, pch = 15) -nrow(notes_MAN_prep) # Nombre d'individus -ncol(notes_MAN_prep) # Nombre de variables -dim(notes_MAN_prep) # On peut également utiliser 'dim' qui renvoit la dimension -library(FactoMineR) -# help(PCA) -# Ne pas oublier de charger la librairie FactoMineR -# Indication : pour afficher les résultats de l'ACP pour tous les individus, utiliser la -# fonction summary en précisant dedans nbind=Inf et nbelements=Inf -res.notes <- PCA(notes_MAN_prep, scale.unit = TRUE) -summary(res.notes, nbind = Inf, nbelements = Inf, nb.dec = 2) -eigen_values <- res.notes$eig -bplot <- barplot( -eigen_values[, 1], -names.arg = 1:nrow(eigen_values), -main = "Eboulis des valeurs propres", -xlab = "Principal Components", -ylab = "Eigenvalues", -col = "lightblue" -) -lines(x = bplot, eigen_values[, 1], type = "b", col = "red") -abline(h=1, col = "darkgray", lty = 5) -coords_man_stats <- res.notes$var$coord["MAN.Stats",] -coords_man_stats[1:2] -contribs <- res.notes$ind$contrib -contrib_moy_ind <- mean(contribs) # 100 * 1/42 -contrib_therese <- res.notes$ind$contrib["Thérèse",3] -contrib_moy_ind -contrib_therese -quali_julien <- res.notes$ind$cos2["Julien", 1:2] -quali_julien -sum(quali_julien * 100) -contrib_moy_ind <- mean(res.notes$ind$contrib) -contrib_moy_ind -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,1], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 3) -mal_representes <- rownames(res.notes$ind$cos2)[rowSums(res.notes$ind$cos2[,1:2]) <= mean(res.notes$ind$cos2[,1:2])] -cat("Les individus mal représentés sont : ", mal_representes) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.notes_sup, choix = "ind", habillage = "Mention") -summary(res.notes_sup, nb.dec = 2, nbelements = Inf, nbind = Inf) -contrib_moy_var <- mean(res.notes_sup$var$contrib) # 100 * 1/14 -contrib_moy_var -var_contrib_axe_2 <- sort(res.notes_sup$var$contrib[,2], decreasing = TRUE) -head(var_contrib_axe_2, 3) -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -mal_representes_moy <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= mean(res.notes_sup$var$cos2[,1:2]) -mal_representes_moy -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -mal_representes_moy <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= mean(res.notes_sup$var$cos2[,1:2])] -mal_representes_moy -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -notes_MAN_prep <- notes_MAN[,-1] -X <- notes_MAN[1:6,]%>%select(c("Probas","Analyse","Anglais","MAN.Stats","Stats.Inférentielles")) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -# View(X) -X <- scale(X,center=TRUE,scale=TRUE) -X -cor_X <- cor(X) -eigen_X <- eigen(cor_X, symmetric = TRUE) -lambda <- eigen_X["values"]$values -vect <- eigen_X["vectors"]$vectors -lambda -inertie_total_1 <- sum(diag(cor_X)) # Inertie est égale à la trace de la matrice de corrélation -inertie_total_1 -inertie_total_2 <- sum(lambda) # Inertie est aussi égale à la somme des valeurs propres -inertie_total_2 -inertie_axes <- (lambda[1] + lambda[2]) / inertie_total_1 # Inertie expliquée par les deux premières composantes principales -inertie_axes -C <- X %*% vect -C[,1:2] -colors <- c('blue', 'red', 'green', 'yellow', 'purple', 'orange') -plot( -C[,1],C[,2], -main="Coordonnées des individus par rapport \n aux deux premières composantes principales", -xlab = "Première composante principale", -ylab = "Deuxieme composante principale", -panel.first = grid(), -col = colors, -pch=15 -) -legend(x = 'topleft', legend = rownames(X), col = colors, pch = 15) -nrow(notes_MAN_prep) # Nombre d'individus -ncol(notes_MAN_prep) # Nombre de variables -dim(notes_MAN_prep) # On peut également utiliser 'dim' qui renvoit la dimension -library(FactoMineR) -# help(PCA) -# Ne pas oublier de charger la librairie FactoMineR -# Indication : pour afficher les résultats de l'ACP pour tous les individus, utiliser la -# fonction summary en précisant dedans nbind=Inf et nbelements=Inf -res.notes <- PCA(notes_MAN_prep, scale.unit = TRUE) -summary(res.notes, nbind = Inf, nbelements = Inf, nb.dec = 2) -eigen_values <- res.notes$eig -bplot <- barplot( -eigen_values[, 1], -names.arg = 1:nrow(eigen_values), -main = "Eboulis des valeurs propres", -xlab = "Principal Components", -ylab = "Eigenvalues", -col = "lightblue" -) -lines(x = bplot, eigen_values[, 1], type = "b", col = "red") -abline(h=1, col = "darkgray", lty = 5) -coords_man_stats <- res.notes$var$coord["MAN.Stats",] -coords_man_stats[1:2] -contribs <- res.notes$ind$contrib -contrib_moy_ind <- mean(contribs) # 100 * 1/42 -contrib_therese <- res.notes$ind$contrib["Thérèse",3] -contrib_moy_ind -contrib_therese -quali_julien <- res.notes$ind$cos2["Julien", 1:2] -quali_julien -sum(quali_julien * 100) -contrib_moy_ind <- mean(res.notes$ind$contrib) -contrib_moy_ind -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,1], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 3) -mal_representes <- rownames(res.notes$ind$cos2)[rowSums(res.notes$ind$cos2[,1:2]) <= mean(res.notes$ind$cos2[,1:2])] -cat("Les individus mal représentés sont : ", mal_representes) -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.notes_sup, choix = "ind", habillage = "Mention") -summary(res.notes_sup, nb.dec = 2, nbelements = Inf, nbind = Inf) -contrib_moy_var <- mean(res.notes_sup$var$contrib) # 100 * 1/14 -contrib_moy_var -var_contrib_axe_2 <- sort(res.notes_sup$var$contrib[,2], decreasing = TRUE) -head(var_contrib_axe_2, 3) -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -mal_representes_moy <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= mean(res.notes_sup$var$cos2[,1:2])] -mal_representes_moy -mal_representes <- rownames(res.notes$ind$cos2)[rowSums(res.notes$ind$cos2[,1:2]) <= mean(res.notes$ind$cos2[,1:2])] -mal_representes -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -knitr::opts_chunk$set(include = FALSE) -rm(list=ls()) -library(dplyr) -notes_MAN <- read.table("notes_MAN.csv", sep=";", dec=",", row.names=1, header=TRUE) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -notes_MAN_prep <- notes_MAN[,-1] -X <- notes_MAN[1:6,]%>%select(c("Probas","Analyse","Anglais","MAN.Stats","Stats.Inférentielles")) -# on prépare le jeu de données en retirant la colonne des Mentions -# qui est une variable catégorielle -# View(X) -X <- scale(X,center=TRUE,scale=TRUE) -X -cor_X <- cor(X) -eigen_X <- eigen(cor_X, symmetric = TRUE) -lambda <- eigen_X["values"]$values -vect <- eigen_X["vectors"]$vectors -lambda -inertie_total_1 <- sum(diag(cor_X)) # Inertie est égale à la trace de la matrice de corrélation -inertie_total_1 -inertie_total_2 <- sum(lambda) # Inertie est aussi égale à la somme des valeurs propres -inertie_total_2 -inertie_axes <- (lambda[1] + lambda[2]) / inertie_total_1 # Inertie expliquée par les deux premières composantes principales -inertie_axes -C <- X %*% vect -C[,1:2] -colors <- c('blue', 'red', 'green', 'yellow', 'purple', 'orange') -plot( -C[,1],C[,2], -main="Coordonnées des individus par rapport \n aux deux premières composantes principales", -xlab = "Première composante principale", -ylab = "Deuxieme composante principale", -panel.first = grid(), -col = colors, -pch=15 -) -legend(x = 'topleft', legend = rownames(X), col = colors, pch = 15) -nrow(notes_MAN_prep) # Nombre d'individus -ncol(notes_MAN_prep) # Nombre de variables -dim(notes_MAN_prep) # On peut également utiliser 'dim' qui renvoit la dimension -library(FactoMineR) -# help(PCA) -# Ne pas oublier de charger la librairie FactoMineR -# Indication : pour afficher les résultats de l'ACP pour tous les individus, utiliser la -# fonction summary en précisant dedans nbind=Inf et nbelements=Inf -res.notes <- PCA(notes_MAN_prep, scale.unit = TRUE) -summary(res.notes, nbind = Inf, nbelements = Inf, nb.dec = 2) -eigen_values <- res.notes$eig -bplot <- barplot( -eigen_values[, 1], -names.arg = 1:nrow(eigen_values), -main = "Eboulis des valeurs propres", -xlab = "Principal Components", -ylab = "Eigenvalues", -col = "lightblue" -) -lines(x = bplot, eigen_values[, 1], type = "b", col = "red") -abline(h=1, col = "darkgray", lty = 5) -coords_man_stats <- res.notes$var$coord["MAN.Stats",] -coords_man_stats[1:2] -contribs <- res.notes$ind$contrib -contrib_moy_ind <- mean(contribs) # 100 * 1/42 -contrib_therese <- res.notes$ind$contrib["Thérèse",3] -contrib_moy_ind -contrib_therese -quali_julien <- res.notes$ind$cos2["Julien", 1:2] -quali_julien -sum(quali_julien * 100) -contrib_moy_ind <- mean(res.notes$ind$contrib) -contrib_moy_ind -indiv_contrib_axe_1 <- sort(res.notes$ind$contrib[,1], decreasing = TRUE) -head(indiv_contrib_axe_1, 3) -indiv_contrib_axe_2 <- sort(res.notes$ind$contrib[,2], decreasing = TRUE) -head(indiv_contrib_axe_2, 3) -mal_representes <- rownames(res.notes$ind$cos2)[rowSums(res.notes$ind$cos2[,1:2]) <= mean(res.notes$ind$cos2[,1:2])] -mal_representes -res.notes_sup <- PCA(notes_MAN, scale.unit = TRUE, quali.sup = c("Mention")) -plot.PCA(res.notes_sup, choix = "ind", habillage = "Mention") -summary(res.notes_sup, nb.dec = 2, nbelements = Inf, nbind = Inf) -contrib_moy_var <- mean(res.notes_sup$var$contrib) # 100 * 1/14 -contrib_moy_var -var_contrib_axe_2 <- sort(res.notes_sup$var$contrib[,2], decreasing = TRUE) -head(var_contrib_axe_2, 3) -mal_representes <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= 0.6] -mal_representes -mal_representes_moy <- rownames(res.notes_sup$var$cos2[,1:2])[rowSums(res.notes_sup$var$cos2[,1:2]) <= mean(res.notes_sup$var$cos2[,1:2])] -mal_representes_moy diff --git a/Analyse Multidimensionnelle/TP1/.RData b/Analyse Multidimensionnelle/TP1/.RData deleted file mode 100644 index d9eca24c87141c07f4122a528d914bd9efbf50c1..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 39809 zcmV(*K;FL}iwFP!000001MK|=SXIliC<;$90!kE7K|xVL1QAhCff^(O7!d>!5hRHu zC5Qx35mXSABq*YiL6YQ*P0m?@Qp z9+?rJK#s!^Z^4HQ7=6^t-E*wHFr?X>Ff0|nKCs;hCF-S4yZxR)QdZUCyAx+2 z_#mfkmWMSar{}StZo`Zna-jPlsF;qiw640GGUmZ(?Ay|PCp%#tolI4AS%f}jm)O?D z;NSQP%Skkao3>zzj=?h}nMW`s!`i9nfq}IKnveA9OF73-QRG3<&$d|8Qfz(@~4=FQ`1K2X;;(^^kqzNclZp=D^QuA{|nq|N?7-B?%MP=k=mpF}^NB z(7)gFhM~-3$hd25A-?4!{P={J`;I?GKW^t=iKh}mjdvv*Z*(f5(hSR`zR4Nuug8lZytHZ=iZn`0=MgZm1u) zB!p%5pgcaGy@xJ;Lz!=yW;;#4LJG}F=pxNwl+_$2|9HD2^3y+ml_Z=Tg#_MJ5tX?H ziTz^ilX;&|sdY?YR_-&%4$@pOSi1&EnF>PQ(^Xrdfmle)ky}BMDJ%8!qncvUia%X-cT{07w5~qpf1*{;7uD)acM15xK>Z@5-pWr`HU-!ED zGS<}>^-t8Nu&%z3>*{;@|1atrcxA0?@<5X!eZ1^+5^)8{&94ayx=Vt(c~B@1 zqaZxd8k`I~oD5%7*vN>#s=?dT%H9jYlJL}b)otjr1?X+@VW4DJ{1;OnCa~p0z~F-{ z>`1Zgo#KOD*pZxbvldC=81KZk`jQUw)?{D z^KI#OF#0~v&MNz2Y>!~d4pWJI=#aCB(EN~t5vzN;E4@?1IK&-F`|g@yGGDVN)<)sP9jcJQ-nzr~BK)xSptCWMr&KXn=;1F@aAF zBNIYh6dFc`rpCIr|G~VfoJqv`Z0U%m3Cu60t@T+jqlp2WV^7V&Oj0kh?(K#cG^sBZ z?)Kh~zCZNi@$|%?r}G<$v-2og<0W@|)L?>-lBJ=Wxa#411^>#s6eXk+5XQ>Shj$J%vcPj~XbZ4`Gt{%SR zlGlp9Yde^=@0)}7j%KR#`QuQM(#dCC6p5z#)HrTGu!L&nhYS~n!ch0l{=D`TH57OK z&K|Gf2^g?is2!+SMxAQCQ%$$ypy)@|Wwy6lP|0T3qiVG>D9<#v#i-aTg|rmExhB$9?M~CpH*j4Y|Ewb`&b-vhF;sxQ_bIu0B24 zOby@luKAiyK0*BnzS~bQnLwxCyMY+&Fskd}x_cg#L8{^>@oAFFP>%WE7V&L?IAe#-bdo7B}Ha(AbDT+arEz+v>_MFgceJ)el?g=VR zVCkjkSwhu0LSK&Gl!p4%Bcg{d&7-QSJkMk2$WWf0hU(T3Gbjinwl*;4M&V8U+j9N( zK+IPAWm0Qx__2qVbbstQ)ZXGCdj{P^wO^DP(u;1w55wnYi9X%@2VI}xf6?_B2#Z*8 zeQk9UGfj5+bA%)Kc0}XZdBS(Ly&`Ro>8XIXDvY)xiE~#&Jn^5qy zuDU56kI?(5&*Rs$y6Wr_TBhoRq{~`nI$B1i?8o>Cy?(QtmZ7?tKEB{_c6q`mL`6tE zN~m*-ype^GsXDu)k+G4!K7k$$o_j4*Q+3`We8-Oy2wpbU#jmAJOn8q7|7Mw5O5N1N z%#c@rpPx_%wY<8Sp_$3gBmx$>9G>GlMg%4T%~tk8O8+Hm=#W)gJ2?wCsCpW(7s}~IBYQl z{r#@FhsY_R*SVcNU!N8#_P&mMu6zedsv=qS?W$nNBdV2*)d5EQg})uzRSZ8WyyOh- zorOA;2;QrEFF^4;+1-?nmY`ZZ__fco2pX+-8_z!hH{V`w-hWolL?K| zyInUCr$9l*2dWGb zww#D8fe)if#A+Y!LoBrb1)WAQ__i!6Y;)I!`w=Fl#PPcz)A6mAQ)wrJay}AHbmD^W zhv93J{p64pt?bodz+>j7o^3_85Ov4o%#+hWkgw>Vy5AujK0c(1SmeG5Z%z#F z5SUa0OFb!CLn}Tou_2=lBi;iR^Bz}>^^Sn+a9c*GCIuWZYn3g&R&Fx&Fz(K6z z)tR8EaA=sfcRf?O4DM$G4sE?d1#WeR5-GPX|9h1BR~)4<<~5NU540;VO56Ei$qrgr zIFcaw(5@6?4`z6T3qB8J#{2dt+`=HEfFv`L#sfMlw0C~D7{C~}dM`^*#YD z(b&=1VmbO1c_^=Vp|{Ff3e!K>kGQ?)#r9=9eWBSMh!LrO?rgL&g05qX&$*)VpkVZ{ zVioN*Y=?5smcFM)u|4;$&ZR!nqIRp5io{vh*T&@rkN|!GR zdTeMi(6jKyV@&RI*Lp*$|nIP-H z{ug@eN*JSMjhmj-A?!r0Uc$p0);FtVmk$IO-BCc zhB3pNNg?E?G3J)9+;y|V7}HX-Ox=zgnCTw;G&1V|whP?XC+p5Z?~Rsaht7SN5Vf^c zh}mwKqNM4i-_wF^v(F(l4N8WDf-60TO2e@|UEwTOV)`-0powEjrL-_Uan-EJvKA)a z_+671o`k;ZwfoGvU&G_o9;5x9CK#uAMhsbZw!#Mp5zZl8~ zV*-K7r7_niU`239daIKw#yiMS^O@l?M#7%e!?GxY@y)ZCQTeW793M1IXvj@4Do^*k zqqfXY^9dR8?Rc_Gd zCop@%C>z^Sj!}hk1eUrezyd8-O~4F2%<3s@v55_WL5n2YI^#8rB#p01b?XpDZl}J$ zu=wnkHIc6U+lv1CtY~$EkpJQ7!9P4QSkKbbH8I8g{rG_7XTyK;h2g5V-R2>A4lniIw^q1dmI$I&~VFjb=Q(m6Vy~`sCv5|K_eml^KPz;7_LA( zW-&|_t;&BI9PrSA;=pfb*&oQOlEoO9RX3BS{>%#9#7iHe7rvp|O)5mg4X>fV zilZ|7AuCMi2it~=$fIc=J19>*1*;uA$JJiap~X9g7>+dspg#FcVNV8v0G-O6AyIUJ z{w?;mRtrZ_$)4O{a{2@`zUR~EG9?chpVrGB=)VJ#nvwGcffZ1xC|7I6LV@8N_`LGt z^3VvY( zRHDcWBQqVFPVkdpL?`zyPuT2+zM+bPP6{qleYsVY=C?AkkO>jWW+}tE7>Ertzg7-o+!RMXdVu z8D?ozf28k#!byD;Akb=Xm*NMSNs_CtKAniBPAQjsu%U(Kof(_*i*~?F`6kf|Ey=KQ zi6cGBEfYEw3wW#BKA=VH`+b(i3FuxGcDC7y&|;!Q76(xUOmlMFRY?>^{h9k$f-DE1 z)RMjYLbwaGEANq{vf+Z_PdXj9I(lI&q%YC%l_zRD`XNimiwCBXD2_-ys6l--S~gzy zZ=soFGkM$|JndK(yJ&}hlW7fzgSj?iav zcU}Q>>A55)n(3kZ=GHf>pSD0L)pZ|RhfbI(=aaRdT}De$W?Mh(_<|<#>zBB@9blTK z?h@;4(qB)Ne>g;M64>7yOvMI`tC5QzR;QtH;V8qvXA=04qwv^6(;J$cZ8-#+jzZo1 z#2xPUqkxK+J@yXAL8A}VwwAOvsPTsIw~MLusM#snI{te;s-VN2Q9r2;UA2STdUah; zou8-15k@Ig5N_dM6?h3^E}Kb`%zl8b8%($5^+r(}PKnKS`w?iPUj)i1P8ev=piq>$ zh3b0gZapljMpb7$YbboC2Y-DE`ya+5f-jNyfyw$4kpki2 z+MmarzsZse`jgq#m~I9_qQLVDk~Ov{Yp>Qq`P3jP?VAx@5LXMA9talLiM&Hv{y-oLu`(GcJ|J~H6)~7fIh`3+CdcR=D|CEq0^};yr zZ@>TDaJ@}L>+q8X5ytd8@7Hwx-X;wH`+_MPz*oh353|~~hX1W%nzah}pFHN$^efaT z?0PNS<@CGUgTza<4ApX-6E!6f#iaDCmNz1@JiP+2D6B@4hwN zYU5wOfkCluB>16{>@=Q?AU?#?BgHQh$dBNk>G3F}_-8ge>v!eDHuD@`)nxkhPwci-$LOh`>Vz9%}KBh=;FuNXA169z6cQpC2A#@DRHG z9EgW6>kxnkJZ;=ZJmA}hi&#$&!$UA0g4Xl>{!WiX8!sRHP)Uln4-$MC(kpoQc?ISr zK0b+;4HrHV zeRUIpvo%Md1E6Os@~m5{-=bvU#pzom$xoUif?EgkB?g&{UVt#O&UNN{KXZe?7yl{(`>^z(2|UcXm*s z|Jbd_H|G6GuRk%!{(Thi|Kd>q6~PdHj&kOIUnJ@Oj{MOY*Nxgx;~s&XpZE8S33vOb zv^3O>P5DeM?rHtq*w?fEMxRgj?^HCMqy!jlahLkc z%!_-szYRDL-rP&(T|dF0(lsx9Yrhi9v0~L5r4>79_n3%WdDP#v!(32P@)M6>wGxd_ zQ`)3W6Gu=9NchrfW=Zq=_jrkAwxa z>sF#^*|^guKXz176F-}3-nWy!`v5g&erIZzAM?_{9bMO*3UjXxp1)6X{=2aht-Z#W zbgSmeWev|cIevHHvg_{sFOEL7GQ7Qi3_<(8^@ z%mOi6T-3WP1~|g0JEi1LuHBCl@x9&u_@c?H+RM3G!&mkDBzg_bemuqQx^0N-rOo$x zcH!PziaT|0)>u$fE*IABbDDjWXmI2@&jpS1A87^jyZY^9I+zN|YrH;+Mw2i-VVg(| zAA3orB6YJ({I-ksuA867b3djq&SUv8V_zoJWSDwa*$xmph;Xr`xsbkaloNZc-+1xO z*hd-4ecZ~S&CZgqV|{OMM9OU=({XfUczte*)t&TlNxcN@<0e)mlXrLeB=Y&13pw7L zR*AdXqq@f|6=Uj+?)ECoEO%BRdmKp3lI7Ua@o{G7j_6A_*3k+-JT=Oan-=VWn#8T`oO#5(rv_O)q$(a@mbB&|y( zs&TvP-kI5MT`G~Kcz@~eJ+7c+;W{N6(JS(1=SgPi8Qr9wJ>?9r6zWq`DHZK9S&U!Q ze81yB>Y~<3wWu}q{Hf>RC%4iCy}1x~NMrQM(qcJo+k|85U0DZi2NI^HDZ`_KHxx9S zTFbgC8i2;06K3Wl*Oa_-9nFV`Z=6X`js7lg=OMdVX_337)OFjDJ*v)Vz_Pn=Tm0|ZzHXFEC zzNKi?r{nPvED8DHwoJ;QHnU^%d6z9^?UK=P{+~FgL!D2QI$eFEk=|(QW8BqdKE7)= zlT?@TLnRh2>?BEF^gZ6!bs`es^-K+kVc9w-1=^(|^{p-Ynk%&r@c75YyynVHI6))A zVxFED>mZses`pzrEJz?p^6iMkU zp*O!9chcs2dO-WwCw+-e`R#!|HarczpZ^15+H_Y$Lvc{Zl8Q)NO_(R<7Hw*&($9dIC z)g7NMzq-yb3E5On2JA$qFK6w4dCP0}$+FXNmnF&d^|fb?cKP11Z0ByI7KWsoe5A-EtqtA-Lh@jy44D`;;diqE~tFld1Oy5b!zB= zS4~c|%`ekCcij?5)Vzxw;PZ)p*j;@#!XP2_RiZ7O;F&$c)0R)=(^NYcCJs3JHgPeO zcG$G4y(dq~_VM4w6`%GdB0tMr<>Gk z54(<$?Kr)ghLYpqfdTBAXWVd?jsD#rx|2x~Ma_BqH=EvGZ&YJ08M%2$#A)-HyH{Dr zJ`0koaOYoK{kCJmD1t7$W@SK=?u;z!7{A2v`HS%nn5@3YDYk?qncg(iPy0tL<|%U(mJoqilx+TUgeDU-;K-j*u_8_=k56NMyF9Q#SxvWm*HUo z>a1gE4SUu;5Oe%ox=OB9KVQ%Jw?UU1&p&@s(P75={^1YysE;jd=`zpqc~jFn>c_`I z+1($N*?3zVm9>dKapAb=oF#;wNRFwOReyU>mT_(8GePT4B8h-EwTg?Y(SFT_4ceQX zUuhom@PB*wsU$iluc)LpW$<7c89jTp*DNo9Q>IgW>F8x&+!l)i{8izHqC;38xZIO@ z+b$2GYD^BCb?l!$OI|o1f8`_FsHLIOE|_40}7Sh zNA&Gb&z;8~d399KpiT3rZLIQdJ_+>^$`R-^78SWk#jc^j+8Q>2Er(#lrP}sNNFnr5 zK2NjO)*`fzz$c+TLb=~N;4WF#M{%!=P@{;PyewQo9jU~dqinDJmJ6YM1U?D%t(W`F z&!GAM*)*jg>Zf>9GXKhxAWwo^2<;>A`MbXLa=-Z*j4L^n-_A+U13{hyxe(e%;FC}v zq1@kh^#9>RoAMX@!xL@F-&+dtOMibP>@G;IoxAk+$KR=!FUb=g&r3@Z9_7T({YE%{ z`Tqs_8}~e`F>V_OON8`A05 z(+TAXc#KE?_WK>uuZP>#L+15*2>1j(2 z>vuzcAk_0)js*Ja`Gk-lkKgvOUjMEQd~d9u(9VteBIL8J$9J!Xg#3+q-dNA}4SFK* zMW~+;C$w*)U2N1Vq1^<2*X{22xJ9#`Pq`j$TBlFBj!&~5(yr5^-Vjo(hXnl1>*Wdg zlpE;NZRppqbnEGad;&g!4x!$^ zs)u-;&c69BU-W~G z_VYWwAtc+tH`Rvv33B)k^l^mW-^MMN{Z;OFIQ@%0%>P3Aoxb#|{O|b2d_tTMF8t!_ zcmBkBd}^Hz0gq7M#(D_(8{^{}(pT5Z5$Yp^8~OXI{J&}kfzOTnZIllozA+@!L*Q>C zAAcdozpr=K^+uS#>DJ?fbs8a`fKSjrA%Ek1Kp2y+R7i1l>B`f%g=d}IE`b<6Md<9hn<_4bDN@AV`> zUj#k=MLRaG!dlB;rF_SkWOgl@&-J@Is*5P=@HU@uRGV* zhyS4-;y>%@_c{`{fj*(!27dnoUxa*ud{;NnBibPMb-Dhom*g+nOK8`|_N;CwPx|L} z{D=BC$N`5V`FXzQiTnE{$^YAd_+JD_{x#rFXRGA;S`W1J3Afw`*OSRjv`mdm2zQ65 z9%vbx=xP}0|GuWY>4C1Xj;^7uhWcMzry|!@*D%GS|HoT2glp9_SGDe`>+_x0GSo4> zL%22cD@{>X^LG->Wi8YHfc7ixk7yKE?`di1YMK0}Zg>6d9Q_IY&)VAB-M_QCY3<(^ zga3Uo_<#Dv;D6s>#QjgW!T4MJzdr1!uWKOiA3p53b7yvcPs0)DP#GrSU^RkTIu-|) zB~mDh)?++qkPaCSRZCv7^25FE0fp5rYD|GQT6)Sg4BN@UpWBtLi_w|7^H@l+U=&Bb z2QTbPggJ65THTkUxW9NKisUamiT^TRZ{$K_V!4bveKx~LZvd`E(gqEAlnrk7n1iuBcVslfQqW+B z&I6AxJJ1N@))i^>Z!lWLcx>@@9}I_Ov+v3#hm6}6GBwqXP{a~c#{2Ccv+vIHxw?`<~9`T54uR@04{`VKEC!jczNzR;7 z5JKcrou{N~P+@%Q(HqP(ki=p9%;NGpsB}AUN%3nHeDglV?wRHRKcr*BmVE;t`>aqV z9C!m!Xz2c(Svtt`GJJ7Rh!)wLy701B`~j37OmfdZ(u1lMQM&Y1KPZ}=bm@7ygx;;P zMeyjqK*o;VRo`qTpeXvx3wBZq6!kS|)UZ|`(!(DeJQwc>=)MdVOQ8jsVwD01mlD9) zDJHWlFA3#wejd8aYyB54=)ZcY>c86s{a^ZK8NoQV{DpC(uNw#Jx^bX&?ad6!@RM(B;)4FlktQ*JLx^YbYg>h_IH;(t~#!ip1EY^*q zZ{0Zj*NtO&-8dZ9jbrjJjKg`|ICR#HBXQk0lGlyn(Z95Dz^Lsuq9;>kFf?wTwp36C zeR&o4evCeaQJD(bt)!FC>tXd{%3mJ_?*@pT_*@Pnk;;?;EbcH6b$dt?O@QN4Si9o) zZx9gWNfUgr7)l*w!**^?1-*#RnZ_f#p~P$03zA>K)7Qy7<7W`Omf#55edH5(%cYHp zr7VMEaO2dY##0b1GQeH5dvk9>xYdv~f3d0YXSH=obQ-TT>v2{(Rad4k?hd-(sO#Une$W?M!ngXAJ^BpR98^M*#)~JH^j4^Mk|J*v+48 z&%)~-x=F*m+Tdd2xop((9C*&@=Z2_Vgb(B9pWUatA$Sob#aT8?>eXEOuE-tGHLUJyY+Haywpa#|`!jEp|tb9EM1#+aHQ*gh6~npcP)umj zH1DSHX-qtU-u`t$I(8((qI)Tc8!C+|`#MU_f!j%egvg!cn7EB`()`QK2c0hQR~++^&4xtI&cba^ESB zkyAsz_BQT}nLucr;!uw={f363}W>)n3T z`n_=_u)+fR3g|9V^nF6zxs&5BuUSI-WXZ+k4@&Uuc27d{k!Vy^)5-r4GdUJJNtAm+!hMmm2hvA0Z`%*Zar=L2R20Uzpx>~&Wu$2j2yuPNUw^E|K5$^_gNPS&yrUaF8_YMaOpbX zS^C0d!v2EvlEO8)|Bk_BFATqsbxfCZgb^A=b7m_|XnGpUc#6UqM(KqYoYVMWkn6RR zvY9`$r=E0LG`a?ja*x=8=f)wVS?O~+=>)jRd^mp+rv=5eRJ0}dX+P|Qfq)Fd0;r}I z>)T}I!J9zycrhtZb^NQp8?#nhHqvYi`0SegQ?(aac`+0n;#OW`{mqzynr(g`APRb zHUY0?{+rwRu0iF{5!p1Z7Py)*5ixay6VxijFy{Iu4iiram^@$8N%mA2BP}TlvfY0OSBQYiA2gumJ1l4qD0%!Vj{MMqxDq(ZXb-3^I_~tc@`#)OYDDh3a}HC z#PZ2Aeb|xH)1QmDqnX5%y6V5d%4_N`C%5Hc^clXm)op>W!k+FabbJ?#UtlpT@KS@N_z%@e zy<7j{u<5_L!T)_9Z6f{tvTN)WEqkc=A;W%#=O7xYHmPrD?w@Rw3g? z-}_(OHHuM1bLFc(w!x;T|6RVt7XxCLe)xuLc_5 zBWy>8CUkZ-2Ifitn!0+_mF^5`DRca~x5pQ1FTV&(-Mt6(@8M{bPDq7Go?Da_CWdIT zHJFOsnH2SlG!A_=%z`Cia&8eevA^yhf7s;yAllvY`r>P-oBw8SP{ECA`Jay;oVWw^ zL)V_@^`)Si;JD|?oF7o#HKD%4wKt(YqS4Ol>qn>~r&rqYQ5kY3_nWUA>VYcCS$eJH z2*Wj)IGky*Z{U;wqd=QTVZeuDag3ud>s z_CWo3%wq?69tg_o|3dFigW9+>T8LsS{(7nUfBrEcH16LGlA}#%Oj^KiyDTXTjBL{| zV;X=J$2UC9aRL}F+3O{#so_81;eYGXHs>yiOa6BZU}TV_X=3Tn769>IG_1{~gCIhu zQ{t_W8&u!>q%FA973%%`2CfbMqjmoOr;HO})OObD`~3AcXl`T0|6AYi`Qr$Gii!uDF8&mvS=z*QTVUXBlkc>&J zr*Fr{5}S;$y?+Z0u@f8UoOW*-XoSP@GU9}UVf9-l+P`Lo(5((j<-ao#6icYz1EkDDb{+plvCUZO};z0)wMtHEr zgX16Y2tMbZr`puNEbQkF>TgpI#UJo~`mvnxApKKBxL^LXvv@d<2eUumZNtOQTWG@T z^!`9rBkip{`vzc6uO3*fB!lP%}-Qf9HH~E|E^|zl%Zbx1K#Q% zn(znQ|AAl5_(LQAD?Xd?o7^93`v3Wx8I(U=bANb;jNqB0!6>e8y}k338jL`CP35*@ zkX<;^Yp*H=Z;wlJ5-&Pqa-#3zgWQv`;|(RxCERvlo3BqL^J_>!yPG!Yo~voU^A)G5 zJ&y}xm{M!g(YZ9`KmQB0Kh1dSr$*;m*UBFDf+BSv3^~O>FZK7 zU6r(tj+v@)tr8sayX`l5nK*R_ucYTgjX;XS*CXt%q50glyQUq z8dAhg?GNfYevJ-eBONl|Ca#BJKeosWsU~93PyB_K^(c1e(Vq0~d&e+^OLescrJfL# zcube^m>e`mQIFcL9LLUey?Dt$Glp&J^0=TOk%Hm4#`Ao}5-};Z2OWKBS=i2ty~$;2 zBrtzR)_PAd6Lxt^qjU7>9PHFxMU8W6*Z*&q;(x=5Ke19K@h;Ys)@6B|NwQ1Jwf=^8CeFKCo1#@Qlf+qdMQ@9DCQ)>C$HQ@gnM6qA@Rj1iOfotzmZfbLnM7V>mrp4^ z&LrMWG8(lVGKs$IP4AxU$Rx6_oejQ9iB{E|DefPM$|Od@r+Kc@qE$co{kN_0cp+Q}TD_SeLjxI5XKDfd0eB)TrPa%;>W zlZfs6eI?@`nZ&!r?~ZlqXAJ-el zp_SBDpAPNeOj3h}Ij^Epv}ULlQ}8G~MCJE!{l9oSPCdq-Q z+_v3#I<@@_Dw#r=M1CgvADHjKgm2Hqv-A4Ur^EKmSt$#qnl^LTU$urtrlZx>Ev_(e zjB_U9OB(!W#~NZ?BB8I#O_7dhJ8I*XO(lw(gOM23{2H?_@Iw3JwpZOFDE(M<@i&DT zXcm-{P1kt}9m~aUc0YUqedNI*x;>T1kFj-EAp1NFlPHXjC<{Z~UHkdQJ;RWDwMzsP z?!nair9vK3F34l0@HtW^41FJ10(ISAK&#aw>HRD+Foll1G^9KQ-Pon6sCUcVhz)njTFYuxZGQ9-jiQ0~qRBI4Z z=_solNY<`ZKgWX+U-y4NW!_#|;hOfcOGvy$T`wlSqJgT)e zy#XT%ygewtau(*On)DBxvV)b*H}uTH#TYqB-N7bRJE(sol;!P7ff4Iz-D0yoi{UoM z_&;87#E6;5_qpLhVdndns7u>LU?I>=BkhO`ta6&s4*5C2G`FFNIq93bu>m|U3;-@(F-HK_h^zrMhaHS{JI75Dq-TglSj3i2~3H~2hm**!^kx{ zvcuZCVL@1F7rDcFB)QsP2GOgM|HFKzD3A+h|fP%zY29a0uE0p_)}I z@-AXf>cW{7|pm-=~0DNw1jp`aOj4*y(vSwn@mMOjxeDT!Q+pT#htgDMh2X zU$WmC9)y+b643My#K<1mCM)ahg}L+=p5drUXgkOL^leEW%vw$_7p&02Bz1TRN8Ttj z9*p78-VzBTZaUj}_pL#T!js6F*%(OZ<8D8%Vh35S&egWeBB+~;Ub-;E2Fllj*sL66 zpuP1@l&(KH)KF(9<~B{D+K<|&Z+1zb_I(0&X3Vb8#8x$!U}yl#vDWX(s>LxP!RRVj zRz@u*llp#3%BWa9^`O*C1L)zg8VuAafzD~#g@(#JG(1MN*yYj$RmU#xunyyaW@)-l zjFHiq#BtAwZqWs3J^9`**)|s@PrLC~^u(dw%!wc_<0zOIITJJ2FoS+1 zOO;vAWx?3kxyxlCU0~Zz?z&fk9fm$WR5)><3c4t^KA@!ugs$X-rZP?ew9K$+u%j^z z6-PRn%I+ROoy)wlsSYQgKizGO$+8Bm1UN+nXyt=Rm%JD_8 z1BRjn*=Q7Mve2@~CBN_D*{GoIm{5DuGHPb=9WuC=1MM=HO@{ajRK))6aPw9oG@#VL z3Zw@lskIRpl%R-ZvQpZn4q?MLOcYGw3H1U|DuH&gIS`64sAox zr?N_VYW6`T&05;t0Ui`rKr>2m`WDm*9viSLTtcNog2&z_N2Bp^+or0HWHi4kN&FH1IsX4#Vq zEIq7IDa%)bx?37(MrvT-3r8uM7z&_xGVcrT?e+Pv#yx1vkLZCWM^Yy4fcD`4k=Ykn#j?8F2?$+rP(g&!EyF_%_EFOj$AC=?< z52D_iobops=%8b1MN3J(425pVs(CP23Dun>Ul(*OVZdi9E7daq%`Xj_CX;kSW2bV) zSHl~qH+RQs4$W>@O*d1!ooN8QJhX4^_Jp9umA#$2YQ3ScX4%?V$sbL=`*iy_$s^Q! zk=J^RC>eSJO833Bp@&M9oq@EY=P*)Js!FACT5vnYb4j+WA1z(15i_H*gONQ>PEHq{ zQKxKotnH8@G+MA+ugHo+t+lJk>DKWueu8{YnNJvu?{3m|9leZdp!|IqHwQ#NTsn1p z?|D?PWK!s@8x0++=XO<=YeVHbwshXs{TLZ>LVK^E0LpJnNTz2qMb${`MH^i{8km*6 zueLb~hVG754(Rg21WW&uD!K#Ev}uR7k1!Xw`j5sZezArg|9a;D8&2p7oO+Zk?F_y2 zIA817zTa{j%1(7T9$f(AMOr)z5{2+1ck^v!Z*DXk6Z*2Xa}!hv8x&#&kkfd5v%kqc^T-YKE2Z$En!$)^- zuhY~}`%vkJUdLOM7>}MW586StYy0?3h6Ge88Z)-N(-Kt)FbL;*yntlf2bRmv)==5& z$Bk4vU!j2Ws%3kP7;2RwcB0t24?at%(pM}=A+xS%`LES1C^$f)=2@yOs-F?H(O=+$ z%tgx8fLK1%IBRxX_uEm(ePGU+GBFFa68gJ#EVP1O=p(Uh7B=ur@cSm|U5n`Zo6koY zS|}iK=TsPVa2G0mIBj_HZgD6hnoE;`Z~l{QIJRU2w$$nRi%?M-p1qrY3`O6J zx1qcr2a&R&sg~8}Aggh!hvo7;RB-A*p=skql;ipEj;?kV>gneYeflU8C9|j~Q<;UL zj#PcexiT)~6|$l^)69Yz9+9RU(BT49LTd6UgAsD*4ThGEe}Rg=VQ<+5KGelnm5bOm zL5J9vv8@3DC}estZqP z#-uA2J0Fg61VTRup1y&~hYo8WGxtF`$^*6BmIonq=4;34c|qtomr{^Mbr_oK`Onp8 z+&~TTE_bhM8KPctd#D!Bhw3jPY@L=tkl5ueW!C2h$^6&4Pae-g9t*_wLe15v$lm)g zi!d>yZDV$S8x#(inX3}ZFAqYdtx9%y=mg}f7WZfMC!(s`$w%FC%%SksmfUwM=8$Bs z`66=k4vaJJR{LDS4efIbcWV{LQPILS1G62;kULRRoj#@v4ObQ->{_f*r~m1=?Mrnq z=HbWhJQ$6-8w(z2H$|fcA4dZ~D#Z zG6i(HYR_l^4~)|`I2Ipsg>FM)sw~D52sp?eMgv<>rmCJ1)1?on_o?OXp+YuPwfidV zx0w_u-`xM*i`$?rM=mXI<6Yv~^ z0=lhbTPE0Gr1e&;IKUJ_uJ0 zQUQNWi+8V96Oa`>Z)@GBJK$dL5CvaIG_?-QKrvS>m*QIOkv+9!pRxZAl;B3P|NXA3kd}Okyq4_-ghm|S zQNqTK%HtW{TPp8`GDUCI?>yxY`t`Nz8UI4`F`0jPL9zwqeaainq2NVDdrrq56L<+( z50|)V3yt8}MEZ)3%Tr|ClJWWKNHj{*5NGnF5<}I_99mne4?zq~@$6o`S13Phsm;Je z7jhId4|697qpa!c{ZphTA;kW#KUu~N6 zHPK3By0uz~rDO&@JMT8mVR9cu*PWPtMW>2VD>v;qDrEuX-{0Jd*)Md-7NX!=FA0EPSsJNR~Jz*#-n#DLk*=;QMG-VXG4K9Nm&0;e;g=gNhk z^3u*gyOKaiXN(Z=n6W_tn`y0Q&W)n1XEn!#5(*%@*8kXnyOhXQRB${ni5B1gjNjNl zPlW6~4Y@t7X2=p1YhKyRgFL8Irxvf8LYhso4o&zOC^Ndk)e>?aG8axJh3K9Kk83x) zyO=kjBCj7NpAJ5TD3XL%XX(1&i;yJu)=#_e^4-%}MS289UC_k!Zytei-o#|qBT?WI zZDZ}Bz6dGP3I^^XM36;w^vjvetWf@NPxwc293<|0^WwTh0ZOe);W=2j52g8JxNa#` zLy;|kW!;}0QFcbki{b!glydu4sim<5stVK+t@)gS3SZBe6};IBIUUWJ7n0iHHI zwt_n-WX~*-Bx-}`uhBPncjHjrV*?RAaTnyHH#1DzyaJi$->gY}ibPq1dGZ?dE|d`h|IZ01WMe7GstXG zkeO*~UB;PXsBoD}$!7XKs`#qV8aR#Lf6>}L56>m~Eys{Oxi_94GDjs^wU>%r!qJa` zw~o2bi%{ud@nhm|JkYnNYXe)8jgfbj`2bO=I^?!xL{Ym{{)0n3&%Zd-8^sZB1^hY| zqt(>X)-pEM(!6D2YOHIhvmu-Gwz`SddfH~(FXtQGuOqbIX@u9Pf5r)})}Ls8dtc({W5^$P0G*G}bezg}DYd4uJr3;yT7 za`@@&QpUp#JgDH|CLUDra0?G=cu>dvdUf^Zzk~Ss+Uif8{ro(~PksJ8b=Aeg&)0AD z@Su2i?$`U8Hh6f9 z`}G>_Q{1mB0?+VZhX;E+JjcTeJiNq%1Mb)LfmgU+uY~@*67U)iZ*ad}7j?yh8y|5L~yCT@!&)-j*qr?3lpTQF=!RU*9MdUl~!mz8C?^lN+7?Zz{a-TU7M*7Qc zA8!|ek@Au4C5IZIXZMTHLaS{srjh#e!sbC3^SgO%$K4?qm_NvC>|z0<{Dr!*UG*>! zY_v_qrU}L}j6KyH=V918Adq>97lzro8>$y0p_8d&_uHIKXmpaOta{cCJ)`}@7ptsc z^lrb@jS5fb=GvNYG3Yf6^ndQV_pJd&z*X{t(gzq~PLhKcpI|guxY(M~62{^f_qR0( z!I%Q+?Mc}!Fnnvt?g$$Zj6Dp}^pn(tk?FxFkM38)Xesa9X|t;^CfL5~p+f|WN{$$d z9!`T1nU8#|#m}K9u*bDZb^tmlRgO(}IKZg#Io>-z7GO+Rr1(Nd3JhH@5=}n-5=J59 zBqIwBhNgUW6i`^f*yTN@zG0hTG{4oe_D(YVxOX&W#nc8yAIiFn9aw~sYgqgE$~^Qt zT&~zOdjN*5sBjGhx~Q9H!To$uyu=@rUML-jfc|83qoaHY>xi?2WZ`Y;Q4BR1w@~Fa(778 z2YNq>UcEw=2HOv0K1*%!0hPm_dK17124WsQ+W$TZ;@KmTE)AK$u=2277x5J^j@1iA ztI5zs{2<_d(g;*_lj)bes^gsL5q_ya#J+`+Hiq>u|{Y@WVpQNJz+6y>)Rc8+r$J=Cs)zg80p%A9*?* zA@4xaz2@_F(7Se$+Nk~|e4igyIXpxLpEnB(phhJKK4d)Z7nThX{Eh7Blplx$wEh!qZas1f+;Fp8YZ532_&~3iKtVpu|9G-^-r2&>yFDD!#iH;?4!m zL}F2pBi|jPP38|h)P*~m@OILDImTGGX&a0bd*AS7V}wz(`A&4xAikf}U83J_2*c-_ z#5(nG5TGO9^P%!0wC>!QZ>jkPMtwp*xFc?; zJ2%y$Ya<7PA~%`0OR+;6!*|W4h!ZeeOZRx$g%pM?Nl+yAT%0+6eBjn+ z;+=!x2X!r;FXW)hZgW#``&$?$-e%zypAQ`#Up)c~&%!Wie7toU5e#}J4f@_qf}z>H z()+9;VJMSNXYITh45kk4*5Y^jEyqC%7Z;)j26+BTCiG>m!0@b&oKsmk^reO-G&?pJ!XD z4a7X$uizEl3AKpmlryIv2piJh9yoOz?1bhgXKzkILBgKn{6=*UEG-bCZ+#XDDnnDO z`U~OiJK?PM`CX74y-)2fS07}oUf8?qSPnFF+-jJU$3bV(>+LiAcJRGoR$TH3ew!c~ zuF%T;5%C2!OY{`U5jWBfwL@>GG`|!Voq{et(}6 z8GK<9?2V@11T{AeDqNV~LQ}`S$z%D~A!LUbWj?P3Jk?D~$T-3ei9JW6KFmLdFpom| zP0ZKe+dReFcba?Pb$$Y_)LkC@)xPP{y?zB<3QLpSyJW$Sn|l07)H?{@6DRj%TMv|O zin?UxR|=`!d*5xP;ediIZO$id55W7jCA!UzpFZL}{=VsH%Q|feXny z8p<1>DDlJHA_qH2wBlEwx!?)|ubBG|osNPo`pgG9FY*0<;Kty|w{?(hW9G2j@e%6D z_tiZpxCf-2v4RV;lH^F zQmKQe#z(YagkL>BRQw5ySr2?jI?)1SVx;*JM{dLLVZ+?8#T=--9zcIgV>4tr4J7dt zSV8y8$XqU*9psnjCCF~ifR3l$&5u3E;rk_Q?yUqFj9Pdw=?WZ!VYM|K*1{H03OYW& zxuy<=!tT)-$Q^^x!448pGesE6iZ)ct%ZKWv^VRRu5}?cLj?nFKDu{UL=D4^&28JU@ zmz5RRU{JPPrkOekvcI)G-S=(|%Df#9?TtSKQ8NmY@4*Z%XU#G^ISF~9W1Zb)Cg4`aym0(F8DyN>MIQs|P^p=pT<)g{MX@Ij z8Jna+^yr<%{f7p@;KU=j&vM0JpZ-)|u7Mt6{fLh|jvEDgro5MzWo4lwwC__ZX{~^EX1!Bf0>#*50zYvv5swFkVtRT<*us;0-ecvle&eF zQOP#!<*N?yJ`9v0I6g>cyRMeJE#kKvbM~%;)2{7+BBPQgBf3KnPd3ds1__Di0t0k-3Zo~IsZ0Jq_9_hV{oj6)*e zN^Nr{;x<5axYPSZI}Iq2(`)P$(gMM&DVz4qC_%os#K0TvDaejAXYTA1fhYFy@uwF% zA!j;&RaEvg6dfpzWD>E24%Zo*kNvTbk$CQb^fgroo2Oq&>^KjucLnYCOFW0V>uQHC zZ^1$Kn<)CphcOUtes%9w6Jf}Rx3Rrvmj%@)3Lgkx$%EpF8f#0D38+}}<7u`02!6Z+ z<@-fz;5EymM$USBNH`s|oo?h5*t#`q^|4QYm)C{JTwN9LqrZQFwviUBJ8Oe|D?dOl z>&v?&7q&vd)!MMf3Jj1cxaHdEJ3Wx4Bz9=o>;iObIuo%bECu(2oKAEpl7dC!fskQ_ z<8V1J@wy9s{O~7vsL%U_6tvPCZu;E{1&@OKLveQAkRxK&Gn{`LQYtUTAK++&yh$

o*gB#8KW{)xad`30hzf1Wn zaDQs)U5Zu)E~yiB7KY=HDsumNnNkc?7t!c#?qh^+dhd1K2sMJ#_3xI;DMp}hr(Tou zVH^bSuol10&joL0m%6?8zJ%wphmH({1VY`J(mT0j%MdTXcm3|A17f2lW79{YvcNNC z;YwnJH~2O5qzQ(8gy%n$4vxN%h7>N|6KE+GJfq0p{;1D`Z-=T1$R9C4e2)6u7KOXu zK4pDEno}A4LN{sNia7}H;`JI>&Kbdrd!lKoySIREbK!k<%VH>!Qwjf^MFTE;D@SFk z1R>A6;(E|G86Y!uyip}Y27&a#D;X~r;2x2s;=^wuzvbv988kV2>mkHi%@58En1QQS zILXv>0r2>Aa*C-N!mF!a%0k!bpv+drO{!83%4i=>JYXyXR;!(o+WI4)K4f({GHeF+ zSciMxpMDSLj^fTVUHJq&p8Pd3g?wVoblWWw%)bLK>*cLyy+&ZSmFHWuLljuKG#j^C z6@#Fkh?aJ;18kyqZ?kz*3wN{9G?bcz#A1dXVJ}9Ef%2st*Sqv_Fg|cseEDq(@NlRe z|BQP9rVcrm*KDtV3Wq++0I5Dm-^#7#49x)^Cr52wx+h|5N4&@9%8!E7$=xfrln%ot zvEaSNT>fB9Hc$QdWHvZIxxIW^;v5lr-I>lwuqTkwoxSeQO%4@jSY zHs&j9q&{Pjr4@#{dpaKCCBcZMX#_7A8{)rw0MM z5$DkO<~U%tE>QzS)pJ z8|_D<13+np zTrurN0a$JyUKOckf)8i;lVq-t0jbSSt_U3tNENzG-!J6@+WR6Z7subhv-wJcn!AE< zR_B;&Ygh-|roJv22!TO18?ukgy%amFzNi8a3OG_`fAfAFdIu?ityYAS<1<4zMq4DT$lV5 zJFgdnJsW)ZAOXL=r9LQesGtr=ih?e8lNrIb;<`OqPtU-srHrDs6%}Z8-~7^}QXOL7 z_ep5aBm3y;VDVX7eABmu~!)8!2j7b=O-)}-4yJ);?WF}TC8xl#d?IHxlM=$1^;RxKfBP-Rlb(^xiEX3s zWZX_nQIZ`$u01Z552lCLObSZk*&JZr|2aE?pArs4ri$prtAHmLd(^k%d;assy<6Ac zh+rnX8S(G4+<%|t{%f4&PJSzvbhVs-u@LQx7-k342K8HyJdA}Yxo#Q%V{epoVzJp2O3U~TTCYB4@nG(ogL6Jz=Bqt0e!msqt>x9B&Ja(+Okpw0xO0M3Yq=$)@&ARfn(lB_P zH`!#r8cfsA#N6C_9tMlvbnGoJgo!dei#LV<%N(Dc3yX6=Kd;*L#^=vql7nn_?!o;q z)IiDY=G_aUnY}t<^vp29Pv>*a`T>lmv586W62W8^muW+U8H{IKpZFMH1pW45tgaey zFu&58xV6p&W?CbHuNt$$xb#Ef-9;KOQJEk$Xk`zhZmn`e!m==@Bu9OfE)8bf1#V|2 zr@@3^*ki8Rb1=2@Nk}T`aTpN!enh)m5$0;u4dOmn!0<73tMp4-VLWDA8QOah#x9?3 zU}`@I(}j~)+BhrJm`3b^_neAWS+m99+P=DF5Ky&=#{14y2z#*-m2&D5)JcWj zy}XwNlJwFT48v$3z3rl%xe*iOUzZ6zbN3k3GgT>BtByhABH1&RUHCQb4|sb+#0`Wc zRYLgRJLA`&e&TIL6~J!(W5>m+cz7jC)Sfz$0+nX9nd9F?;l)F$k;NBH5aiILQRiI_ zl}v}LD~R2oqW9B%mh3r@+x^j1-ZTr`N^#s>wbS4xRB9e_`7$I{Bm|oVIe_=0Xr7KO zQt;_3F0zE!AKa}-bo3Qj;g%?I@k?$Rcqt)qSX%lx#2GyF;)vTL-1Fv(X>V!@!NVYxL4{afIBf_NRFk^RkFm~u6MoJNw@=?_k>aRj%l9s*J@bVg&d8nit8y?j#O-YJ0S7gQSJbLz z+@Rc&Ci7;*N$68cS^av;2?jJSj1sv9!OYhMp;+Nrn7W%cpyr_m11iY^$;o9AZaX&9F{u9Xfp}}?#D9x7KNdsJAHFGH4ig z7(kQrri;BQ%22$NGPge<9rB*>43J+r4zla}2FvsQ!%QV<(`&cWmG+!J>ne19I}BIB~Ws@HzG5sqxp|P-953Cv*HORI2eCX`i=( zE~bh)t2=ukZ0!-e;vR==A^ykb$s#e@j$JOB%ts*4U{a%^vKMl{4=9TW(ZQ&REs7s< zfcp1s?&qy~VdQ*?cq`3&NNd{d;h-x7ePThU1V&?lPBdW0;dB1bp`OS}l0E>HPj~AM zEXzYxYUiOPsUnPM=F1AVeGm8wmu<`G#_PxMy zk9dv0dyxn}h0hOtrtgK8bDxYO@%?`_k;CHTUi|!3#@#=C<`!h!G%NNR`U;s63PIN{ zIYYO6`Tlw#PRKl*8FS4i4DOKoOj!}}L5TITvb16y+^>n3zV>bc;%z#0f=|^!04*9j ze!C6^xn(H5b%J2o$}M!$jm@y!#E}FlXQBD%)sqiBMW8whw@u5J1EPkPU_x#Y+N1gG zs+Y~7=>C_Ry<%f9D@RXOV*U;MRu4(t-82o;pHG-DWfQ~58U4E@O;I2ky@c}#;)LF^ zQ|~E!95IUGCwsRBxk1M^#r?CQweZgVSctXq0Z0oII`D~B6I=(ky|OKzho#DGui?~L zsLXIad`~1EI-;6)6o`~NM3fVVPr3Fq>*M@LY?-+Cz9vckm=2{?MSX5WZoH| zCs}#~V=t~cS#2(dY>Dwq!}bSIt5i*L`sO|e)&92jE_DfGq@R@;lz9MrCwz0WJEt(* znq=B>x~I^kwDc}}xCYAY#ft~T`~Xe9_Os12YUG}&=-G^l}IEV zVq>Vvp4!X8os2`dTiynOx6nwx1gSHW*;$K5^r^#_otyG8GgXXfw{L)Q_BeP|+&{Zn zs0O0R?%KLji$QBjhls!JkxoIHyu4@N)-YFrNcd$3~V3JhuMoiB8+Wbf=06}cYNk3^lec{=o&r%F=Gigb29iLr94*ALp~0w>C{Kt52nFF z?ZL5TRauzX>32ns-vaDZ&thHiI#6i3?ac9kT#W3JQ)2Ll4ZIx*c&^Cu3i4S_OChyd zm^6UGC6h}q@l4lhM+yxz+Q*NtS<}FHKkPdeDGgbwu>(fpfw1IjyH)qbE~vSKoR`ZM zp)0lLlF@N9_>pIL{u>i53=d&W`(zHn_$f1U-n`EkiI9N#4C00F1s7g6*z5(reK|Ag zyB|VJ46zBPr7EOyO&u){M^HKIYtDXG98$e@l05Big)b#0Y>!yjG2B+^YW6tb9Cz)L`T8Hf{VUqWwYvI+IpA(#&mOT6A_ zjM2xwn)x8H9adW-_3O#cLa)m^0nJM;(7W(h@=Tj1)H`)D=oowlO`=cqYNC3eTZtNZ z8SP;HVf4{PD|L+FQr2b_Uv*gdJaJv-RU25+cN2@cQb8Q)N7--rV;Ipwnaj1@D^OFx zmoa_83@Ryv-fq7xh*1t7y~$qx0i#NCpYAGE0*+U%A$#!alZKwVJGZdKz(BxaIvt}w z#QEQ!>K#skc8}8soc0Pp?fIh(TLyD5Vjt(_`tI8p3AXR?x$&zo5l0HRO=>aI_RYfC2GOBQ95u!9hM^pP3v2cc>*_(;SA{#*|O_y5->Q;G1mk*biAlj&77Q z84x6Clby%m0E*N&p~+-Gfv3OfV9EB_Md78P9xJ0)MG_vtv;nwYWG z|LBFz>&`Xh?>T|1+$+D>SPx!aJF5Fkt_!A~e>=R-{{$pZQG3nM*udnCME7GAMNrkC zuH#MW4c{^wTf@(nK*%HY_-E}FU}Vg_dXhQ{G9*>J@AMo72V+K`Ri-o0_uTcB!p-*> zaZ#_^VssaLl}aq$REVFSTHS&Vn$&|{-5v`Tv-eO24DWKIa^S~)0T#-auOW`3kVK4jl<*>2MfBkD;8ViNK&)%al`;leiP;(s3* zn-vebxH9<=i(pXZ=aZ0L{Rs8$WXI3m@xjRa`l5A5!XaC_rp4-KKi_{iN>$Q*j7x8&8;z(+CA*Me9E>P|si&^+xpyoX0>$`3sSQz3rq5kB z`fLG2jLqO3Gk-04Q!YH5-Ah^BdkMdP>lza3TY#986A?+&3@|i5YI{+X8GhKDJ}dJ5 zG1Pr!RZ2eL1CyN*?kzWjVJ2>1IQYB{ES{dx(Ye0~5~D$mp*xbn>1wA38vI%~E?fTbRK&!0OwPPqu01B_Lfx_2?+X1N?E{&47hpxq#I z>Hu7`?jF_JISdYBiCSl}wV^&+%htsB5G_8n|yfX6iIxT3DnZZmG(A^Sc-b-*O;;NSguDo0@4o=x{A6>ah z0AxMUN>?Dc4kI+#d=7`x;k}03k>uO-&>q`5eJQUEx({hn7m%?6x)>UKhHnz)ow%mt zaaPcNlwZ#XzJms08}sa{fzJ(=k_z{V@axryo7b*6g8iWfbcIK)VUEGtqvez+)Vv>P zo=nMri4HaCUFmPJUG*c~14<(px#O{mp=~S>*i{q7x=8`%dT!bapFaZU$Hfy~m-oWd zCzlGAcWbcbe8*Pt?MaMP`$63FN#@^jjIRFFBKj^8=DZ$|s~w)h7@yt~@0>pk^K){i z7OvPq3#RjJ^&1U%K28#>J$(a)G)yHJ%qjkTmizCsT-?9Ua)15XW@4w$+X{m^FrrV^ z{j{QH7zvxrqgwVhjD&pLCA3);BdOTy^)00eBfZBWCii0xMv_tf*hQ5JBbs|QdPc$n zBTaJ(Ws&s9h*o?I>2t1QBscVLKGz7wh`-)0Y4Kphi1VHYO3=(>q_uvE*^bH>je5M$Y_ji%|b@yS_Pe+B`d;r^Il-W2=Zi!LR2i9uESYbr+ z6V!a_A26bh%-dfRA7DhW);@CnFEG-Q1jV+5dl+fy=1uVyF&L2(^>}P{F-9Exqv}Pw zBSyJJ`U}fi5=J`q`jY7yH(?@m*Rv5iBM?!6s19R>2h4-`LV7czcJ-UK6 zm`n+I!L_m-`eg?_9+if{gboLNV304Y4t#J=E%AbhFS~kLyvSklOXSY(B*4hTIoW$O z$Dz%lPgwTV42Ij}#>G1t1^o}UCxpwC|Yp$-_2QA@v;Hvqko#j3B5M8SfIs0k)D14DZG(J`VzF!*hu z=5?0GpF!6ueJ=E#o@n#i z6$iDlj2%yEQlQOhN^Nh519Y!mlRQLs6c$;|84z*XK+oQ6UCv%oj9B0xJ(b6Gn8U7r zuoxr3h|iV{=Soj_BQc6;W#0-0{GMdy3lpm*BP_n`xNddP09NNCk$spJtl)N?!q8(_336XB=hTB0 zznNgKSzDN=taLh(x{q>VcQz_LHwS zjSXYCh<*dXr$iX({Vtj=%^Mia*_%e{rChN3OgE`6XB$lR#@JQ7ABTDWfj3mKr5MHN zxAril1&oM9bn##TGt7|*KUutS0l&vgeEEu-5!T`q;;NePc7F-i&wh3rh6{ROIIi;X zw;YKktt(rj$uZ&@%KbU?>9A&Y)ymXm5LOsA#k?@($IA&f8o6@?)^@mm;c%OVi45Ya zX7b5@_!`ZBW^Ug23eJ|F9~e{DFukQ~XlkN$|1T(P#{K;}btKsB%)br2|KL-4tN&nh z{_py}Fexl47(e0Y>x5CS>5+yRb@*ZToaS5*FJ$ZU&En|z30^uhbFAld^zddh^+;Jw z)MNn7a@sq+)P9F%9B8!0vq;da`l$cFb8a+s&kdb=DT8La8u~Ojh|u(ao%9pdq6SH4&~yRr+`{2rH2>w0VT0W#H0||G)-JCM z%^V(CO_riT(-G6J{hw5z>1UsZKHo7$6NQ_7GjX11X7XwKy^3HoF;aEkGx$214tGzg z9Mnb&OVtqzYR+ha(o(Ei!x>FCb#=5`ccaNOg#&Hq6q-8Lp*ss&X!?@A)S;OxXnIrA z;KM5sXhukcf9cQ>G)*l~lkGZ%CJvYf9N75<%|8mh=aFNIW(RO#!-+O%Dtjc6 zn+nYhh6ul)T||>wYfsg^PoZg&#)SE+A!zEhnR^}A44PGlDhTyhL9_a0AxiB-Xu9L( z<0>TQ5HTQ(&ca&UR~TDT;okxGDUq094wV&71bD zLG6|*pc8s0KXwZ% zl0tH#OUF@afF}RcNYNT+sH(5N-4?zJ`gZYFkQ!cv`UGA7yLZ~4QLag*HtsOAt6lFb zvtofh2Z_l?wQMkOzug558o`gg%z7u`7w8>_XZ6kfHz4l4vundQZg66=-_FY^4%sYY z%3U;i(A9o4U(vk-G7H&lQ%mndC1bE)tbYr11@=};1RjGfmjZ?dmKX#!Z6!US<^p-5 zH)VMTpFnQ^;!|zAaH!WB2G&R2P;`7jo0Ix0D%nd;QJg3bd7oAG9W!Z$aCL{t7rD`p zMf*L=ZzmtrdezlF8@rDx^Av?XX$V4&$h9}8g_U9Av^(i`^BvF=eI;sJemzWwXcEnO z?1RCDmd4J;9GI}V;Ch5S8W!>ClQ7oCHRHyCx?;3(p9slIT(7= z;+AGz39~7>eQFl~)`*Hqt2!@1Uj!XLU8gBbZ4KWxJ*WtsSAx-&!;P@C*Ic-ag%SD} z4yGMlrh&00{gGAt9RK}R{YHTeBXQ01NpFc{PtY&T`m}T{+wl> zKnM&76*tP9m4^7J^h1?! z)a#<=sV;aJ)paQ*?UB|(U8^-sS(ps;1V7XoB;|zpVWz7#H?F{!XT9<6Jw#B}F0oug zejiPSX^CR>il|M?ogz`I9wyAKc(?lHK~@#7mb36_)UT1oEKS*n#v3PRpcE!RV+y z&2cwbXpKr93RWtF#Qss~F_~J(R=X5sy8RRC9*#aa^pz=v(Lr92K`=cndXi z5i*b1Xpl%U;Gc>DADC7tQz$)o+)yXs-jA)Iw{Bgl&)LQN`^*!M-+@!qVhg@$QbUgVN>hk$U;YG>nc`GhAJS1n(_zo{G#eO? zdu`0B^&FmkWOXL@O#9##<=%p_0BQPj5?L zsH^FNZl??{475d?q$LfZHQWvT_N+$mlVE!qznG0CE>guy=(}Z-AJaQ@QHl;lPnKJ& z72Bbem)=3L({!kCG|lsVZZ~StIEy>wr3D?EW*_gN$#zbXA`MLL@rnaQr`EK1Pda;auM;Rkj z7Ipvd@XJT>U&}hxBwj-uZ{|5lpBhvS4(46;rbF{cmB(dEk2`yfpTJ3DH zL4zc#$;sdDqaLAsDhI6m(RWEdX0`_%FnqzphhxeCIxH}zl9gICw{OpDwlEn~rlH8I z)iH`@LAdEl9~YYZ=CDbI>Kb^*gZe5dTzWQ@ zV>rfb#2HIE(06!h)5Wo4sQF}ed_GeRG}C2g3D+J$Qzj2N9WBUF%Lkd>u;W+ zNHz;qtN~da4Ne%DCJG>5*b45{#?fbtQ_!-DpdpeG!*(&6&{m53Nn9J|Sr6K+WXNP)-<*eE;oX z#W+lK?s~~HJ_5~WP7~*Gy#hDmM_bdMpNC#Q^WL_+KUb3w=y6cV6bb`Yp$i zO@=8RN}e$Bdg@*WerM3oXZs_ln;eajD7{g%Isw%r%y~x}uEY57v+Vsve8`*jv-%C2 zN2pdIvn~9>9n|%%NAGp`8OU%5E1&bOfbu7w<#tS`K@er8>{M(l#F9BoHrWZmFj?xo z^1)GmeK&iTaSvJr zuh^dobwX2_XAVCe{{ivZQZp4J&ybUA&Au~4%_zf(@{KIbBpQ=iIW@SQ7wWiF`f5Ln zpow>uE|0`<(3r6A_SyVA7(z;Uisr76M6>1NghxGmPgFhiw#xtwyl6C^_cDgkCdnOP zF=cAC{f|xFtJV!;eli$ecUZY;G*a&Z^2K2^809$#0NCVJ{=n-nUeX`-51}!9A zahd4R^Fkf8D*4F{`0>NhLBZaU9R`{m1CDj!cVwB*)e1gkqC6G3KHT>;$k`>>I~ZpP z_2-NH7hMr*m1EPKVBC%xT#8d;{NJKwFJID_20>I|{C(7sRTs_eHD}SOmOxSM-1XTM zX=vR4pyIjE6sSSQ9N)Qp;YZhgk;uXs=&rt`{eEFHn(vh=FN+F**{5U9ZeLC#B;_hq zd!HRTI4&>oMXaK=r7xmiWt~y}cx~tIAu?z``#`eeL=c*xa4D_cz6jG#jABUZJkY?& zVj8ueyJ(uTvP|Df6E(NZ?A|GU8Wzw9vsx`Z>gC9uknPGrqxoDbv2PYpzxtR0qX0RY z^c|13kKsg9>9uhumBi7iYF_)bsZFS-$f1~9QVL3=tqO!|-=M*m!msbATT#Ondz|2t zLe$NxMwA_rg&HJj`EmKrpvr5Z$g1TlEERT;_E8qVWVOJ;hlvI>{vxMeJm&yfMl#pR z-P)o52G=(19T$|fmA?s`{s0wjyX&4*9Yf)_UZgNjTtn?4xgN8l6HpbI*G6Hp-tJog+M8qD-yN-i-5t;={sw#|hHbG}! z;UX-C?R;5zy&F3YbfTk9>c-5FlgQ0`zZR33`X*3>V;ph zILy+~@|;zkMR_)N$EfEQQE%@U2}i36DiR5~bAOvPw8*oPDvF+kib)@-X_FWf@^ooQ ztA`&dmDP&4RM{Y}%kmrd&I~B%V%~QB#5e&H8|>s%Wz0YJokS=j+ny!az6?ep-}lvj z*^FGV-o~+SQ;@}6>%U9b2h9Yv1j_F@gX$hSTMeHiLDMTvS(W;(s5{n0D7~TJy`Coovw0NdNG=CneMxd{zve-ia2&9{k|@ zoB|70RquC}H9_sY7rZ0;&Y+@-_yCFR(P%P%B5jl1ESjm|z8T{t0HfDmx4f`Xgc!N2 z6Yf`}&=>k8@riH})K?^RB;nLdCXV%^Y)j4%w0!jULm3fh>T9Qy{NYd-FLt~?$vK6_ z4mVP6fAk4IUtBw~^LP@}CheP{HMNISXG2oA7elCa`n1Ap&Qqv-F|7JVoi6gbALf4X z%@B&t?xc8PZ;M*Q8XvbyFrvP5cf`fsUV+8Dicj3bme6rD_2ce)fly5vB2}^N3KYA2 zyOu*E2}PI6n|M#rp=68ZeIi@tP*r_V3V8qns!{lnqxDh=CgsG=9tz??{bH44R-au_ zp8}tBhYmN)$Z%E0E7rkm^_$6H_FS~U+P3DOXb1JUTe}j2nxWH9<@WxjMKpjhyf|KP z0b-&L^{qCZhgEB`j~Nml>~4!}qgZ4JCCFLA-pEBpQeFQR9`fa|h$M!Q!!e+AmueAbsUB`ALUa=yiM- zI?MD0M&9qSVon->#Rqpn3$C1pE)k-#(*vHU?~>4gyxrlbZ|R8fURQI}Fn%hC=a3#6 zKV9W0{(}hhQrCnCoWhUqf%7)U>awP!f_Ib9LRbK$qajqg;Mh50yGK55CYolI~ z8<&r)iNSQz{+reT^%(WB?TqJm<}-1|BK*hSyQ7LQgKd<{u`o9oz9q8oB=YxnmRJjV zf&QQNo&>I@t^1p9(^Y98O^7H$vnj*ciM)7GL`7&&G^fcc5;7-co|TBqq@GH3ibrK? zFl6X8(mV($q5f{2ebVNBzQ=pt|9$Use!tr~Yp=D}T6^vBwD(!3@sc=HEay<~)}G%u zqQxbZqpi`fSzc7R-~~4OcSFdUQufMvjuKqe z(sqzl7z@v)y_s|0;s});UnP?pbR4L|v%>O!83>h8zm(oFx1);Erl+f3mf;=Q`N6)OvN8)2%&8HLRw~mhJ5K^YI}P1>|B{D2$%fj`FlAQwQ+}nl9^D+7i_0@ zEWa3)9NRDGW$*3og+E@;@wNB#aCPXyWH-g z{=N^<4o!ZNwgNgtXjT`-z6ao+{ht>CAjdwp0`UCC<+!=-yu`DP`lxZm_fBBFk3 ztF48ohJCsl<_c%P2gSH;RCO*@F4V2(t<13h7Dw#2|5aFK(iJ zl5a1MF87BF#7fvuTVbtZ?N=!R@F`vzu zu-k{BVixs;mLA3NL<3!IJaRJywdzd zMK3Ox(N(piKwq2F3!dJwi_Sm37~ZR_EjmA05GrOkRKA-ui2gK}03K6OPlxskN$=xc z`;pXb?sW%A(dqPO3sEmjC(1E*u0q%-=l-LyPx z4+F6rTpk)?enu{xE<@+T<-l^tbV%F%KZXo&5Q$9Blo&^WFqSu!2Vp>eFA!3jGQXgZcJOW2Rgg(0Jt zXm=QwC(4cOBIU>u%npqiE`um=u7mYK18_{X^8om`X%}U*2DM_ z%ctoWVt+90gVx`du$$D6%SqN3ruU)oel)~%vYtsjazs60yRd$Y%#I{}4gi223n{Fp9G$d@4cH6(IWIkLDVFJf|xLh0}owOIr7bGCg$4}IU z0G*#j_@9qX7b5tDiSpt6n4e9w53C=jW1Nvkq+=Y@*|a`vH$P!FhNQnqe_%ScgH4o! zP4JU)Nk5R~;d~_@*}mBX9pj=zyCUV0dPN92Za26*xL(Nm5G3p++Z9%%mc!qt`k&l6|JLE%i2NRS2t|LE#lXQ$@__o6?hTcHq zuV^`#59=fKU^rEMq)A{5Hd&zYgUZ*h65p+6U zACl{Hl1{E$nDrx_&aAfyoLNue`oi_dqa5TqOop%<hLGQjNB+@B77 zulbFH`BcE^;3H z+I)fYW4#zB{X^P|A-0Fc@%)eHAH<^hXg@IP9-NNL`JUjz>j>7D<>7Q@-AS(xKhx8# zYdtaRNERU<=OgU>%q~pF{%s}X@ezKf{mSU&=TTl&sVGsJA z!yWNH*_C5|`nIOXQXd-^M|CeBj}8H%_(e^LPnbSASV6ykiO*BD@4)f>SfTuFe}9A3 z#ed@p<@1*R^|b@9RY#Y1nDM*D{D>Q?h5b>i%a=xY(nLY5*4}Y3zV)@i z&g6G=!{^=jy?ht(@>qwqt~J^X=QAy}fg=Fu) zU=K1yP@`hGCvHj{JPB7=Z00Zz?j3wBD!tbWu7?I(RJ1IB?W@e~ZwEGk%Q*Wtu@^p| zSq}4p#;baxK_41h*M+yDp0|ycB-A$`@u=}L3l6Ix)+2fC1uB|wyJBN#j}4V@yoW|~ zvKlb?=BUUUh&}a2=AMOe`~1C-#S#&=ru!mfQJ4APn)PmE@nM2Ir(!O$XluW|a9mNS zbJ+UqxS|kKzwWLcN?Kho?Jk&37fiPcHlhnQvI{nf=ro*_dfK6A|V5(k6_>jn+aZr6bQ*Bh0 zB~_ib;JD}^d#JrB?yUEFE>(K$Tw(9C5Z`;IChfEcz z=)O_Y{KJAE@zR9a>x=imO~)&D1dXgAUv^O7w4~V(ZE)|*jvJm-9{ck3SINP!%YkE^ zGvpck_5n`Cj_{=tV~m&Uj*)^3m2alWiw8kU@kuKvdP;>Y4B9(3y^6Xmr?0h9Wes(# z`FhPN9bw3|oszW7cREChuh`@0b{QVaxNV3&y%SQNy&C_%R-C$0x>vLQmKG%Bbl)&j zlcKJw7KX_@&V?A?0Hta1>!_R^InTCOW8%O+iHjc&r6O9AvoVBGBeELF#N(`UKgl4E1BR6K?dI6}uGNRe3`4@Q9 zCuQ!i;!#lLap=0pja;aXa?PsPDGinHZDq59n_!!w#UXzU6S)4=E3QX|2t=ATAG&^b zJ)B$nVkuviITWhs$v)X21tIHeLUy@Gz~#Du0_Nv;gROe6smryZVEq*BsYUxFAnK0# zrpm{UA>-7nJPqefVC!IP7$5W&QnzLlgio9V;|@D?TdAuE2R5ef8)-cp)=v(1s8(DF zON)&Xrzf6+kUM5&atlpCGfjV?P4ozeK4T~#;THw-0yx>;b^vkx0%WBJRDzjxl;}9D`j2iLNIm@J{Lo*V&|s4txu^Aypuu+Xud}u^qM_Bs z!ylWUMzT}FqGCtdBDKTI&ghJ8Ktmragn7olA+;f%ZkH%uXkFyIO}ge3(s7(S+Vsd> zBr#cY_oIuuk%IhXB_V5FH0|W;OU_09&@!f9K=Kl6GzdkvXlo6C>_Mi6@=0ZgU7eki z@IVa>^c#D`c&H4TG9*DvZtx^1o?>4MRa+6r>BM9w`a@!oO3yg;JS6wjW9#YxR}i~u z&M$lu`oWu<{+@~R8-}0w30VX?5~XXdGXU0L5>|JQaK~7}OV*FsYp8 z)@dd4)KW!M){s5Y-} z&esTpP#HIa9lsr+o>S*>h&%d!y<4 z27k?LRbxDRR?CL7P?Pj5><#-7l-MawN%K{KI^~5dw{BCRqW8f_;R#~!)cU^KukR%x zuctDytUdx~T&mWVOj-b5`tgpgLz5unT7LZ%ghOoFiRCxvtHG%f zQ=!+5haoV3W!97IdWaP;40lf(15b@|6Jqbjfp@yT+6xtXIDG4QsA~2CINx`*XM9cq zWMo*IUyN{^Q2u)2mAb8la9nj)`HG{d;FxS7lD{Mj(q=Z5N2~IIqk>3Oq=psjU;a|X z<@^WOF8P>qSuzr?iSKoKJ~alMwcPKmYrY41dy3EZn&SuuLmx)WwCe#!kL-2*$)Eyk z<77lp zab-06^sw9MLVb{Suf%b8Lsp_m57e%B*-Iknq7YSsJW)s)GpEO9st}2&_whQ`Fbc^9 zof^J(nl98e^-K_Z`#&*~2qV#Ru=3Tu_ed-;SNBTsPH2--+q5>l0ZBG#Uy98whPHF& z$pr)2;8kMP*((CY(DeG(9kJWY5PzS%{t>>V#4PxqaKc{|dr=7S*QLCj~NG%+Y#CkN7Ww;VcB6AtHk^bZUvJ;yv?HXc7?;*%Uf zy*acz?DXw_7@NiZA%WfBi&3~4+H_b=~g8f3;#<;hMOD#3Q7Zr9}0 zC={)svTukL29F;=6>M;{9d7mp&Nmpm+O<*%GK5d1joM&E)#~Xu%T;fus&|!TwAr7f zGFtKvF5j^f@t>6rQZ}6rukZJ(8&Dh$#Y?-b%DK9ks&-#@$xxIHZ#Ry=CzUmX$}yW0 z>Ej|v6)S3Z^F?t3ZwCM#^o;Bdy=#gy{)xgoh^Qq{W=}P!BDKeF(SO0(q5w7FIF~sH{zLhGB`pRG#I8aobq^ zAg@SziS)%ekT)o%bwQv7q#aXF+_rELavZE{52Y$Z_<#${M zod^wvR6ADEt3ey#e*Tkm(Qp;&XaIUX*|5O0_ydI!p)(~s1Z^@H4; zqnEDihS=;7DqkoYZq&i5eYX|s;Qri>qR|x5)e)aJqtP3nMx16 z6XC)1We%4vsQzmoe^X2Y5xV~fLn^EN`c}xuc|oUtBV>K)4ZxmVb)bbA?Ff=y)d4L! zj?N-uj}B;H(gEdZ`askHM_YH5qbEOGX?d^{63WU77qbvT69=vgoj=wDji070y+w8l z>f+*|72x%1&g&tg(Y@?4{n zkzyHCFUn7Toi*U&EM>CUcZ~wvvbnvb-;p4=oR-tFcg#<4Ty5Z*w4E!^Oy`|%ZC1*l zVVfil=$a=X*@D)vv7SSb=(@tBh?gbMOz9;q2y20InXs6Gmg5Zn?Dv;NGepFZ`R3z| zw^Fr`h1-Ky7t%A3MNV$P{)wW<&8|ICI_m%3rwdboJg xl_Cgb#9`Qumz8caFINyRGw^U4U#A{!=lL