From 283ca8dfd9458ea071d759cc37224c638e49c5c3 Mon Sep 17 00:00:00 2001
From: Joaquin Torres Bravo <joaquintobrw@gmail.com>
Date: Fri, 24 May 2024 11:02:59 +0200
Subject: [PATCH] Ready to test new integration with DT

---
 model_selection/cv_metric_gen.py       | 111 +++++++++++++------------
 model_selection/output_cv_metrics.xlsx | Bin 31899 -> 0 bytes
 2 files changed, 59 insertions(+), 52 deletions(-)
 delete mode 100644 model_selection/output_cv_metrics.xlsx

diff --git a/model_selection/cv_metric_gen.py b/model_selection/cv_metric_gen.py
index 2042928..b5f4a08 100644
--- a/model_selection/cv_metric_gen.py
+++ b/model_selection/cv_metric_gen.py
@@ -175,8 +175,8 @@ if __name__ == "__main__":
     # Metric generation through cv for tuned models3
     # --------------------------------------------------------------------------------------------------------
     scores_sheets = {} # To store score dfs as sheets in the same excel file
-    for i, group in enumerate(['pre', 'post']): # 'post'
-        for j, method in enumerate(['', 'over_', 'under_']): 
+    for i, group in enumerate(['pre']):
+        for j, method in enumerate(['']): 
             # Get train dataset based on group and method
             X_train = data_dic['X_train_' + method + group]
             y_train = data_dic['y_train_' + method + group]
@@ -184,63 +184,65 @@ if __name__ == "__main__":
             models = get_tuned_models(group, method_names[j])
             # Scores df -> one column per cv split, one row for each model-metric
             scores_df = pd.DataFrame(columns=range(1,11), index=[f"{model_name}_{metric_name}" for model_name in models.keys() for metric_name in scorings.keys()])
-            # Create a figure for all models in this group-method
+            # Create a figure with 2 subplots (roc and pr curves) for each model in this group-method
             fig, axes = plt.subplots(len(models), 2, figsize=(10, 8 * len(models)))
             if len(models) == 1:  # Adjustment if there's only one model (axes indexing issue)
                 axes = [axes]
             # Metric generation for each model
             for model_idx, (model_name, model) in enumerate(models.items()):
                 print(f"{group}-{method_names[j]}-{model_name}")
-                # # Retrieve cv scores for our metrics of interest
-                # scores = cross_validate(model, X_train, y_train, scoring=scorings, cv=cv, return_train_score=True, n_jobs=10)
-                # # Save results of each fold
-                # for metric_name in scorings.keys():
-                #     scores_df.loc[model_name + f'_{metric_name}']=list(np.around(np.array(scores[f"test_{metric_name}"]),4)) 
-                mean_fpr = np.linspace(0, 1, 100)
-                tprs, aucs = [], []
-                mean_recall = np.linspace(0, 1, 100)
-                precisions, pr_aucs = [], []
-                cmap = plt.get_cmap('tab10')  # Colormap
-                # Initialize storage for scores for each fold
-                fold_scores = {metric_name: [] for metric_name in scorings.keys()}
-                # Loop through each fold in the cross-validation
-                for fold_idx, (train_idx, test_idx) in enumerate(cv.split(X_train, y_train)):
-                    X_train_fold, X_test_fold = X_train[train_idx], X_train[test_idx]
-                    y_train_fold, y_test_fold = y_train[train_idx], y_train[test_idx]
-                    # Fit the model on the training data
-                    model.fit(X_train_fold, y_train_fold)
-                    # Predict on the test data
-                    if hasattr(model, "decision_function"):
-                        y_score = model.decision_function(X_test_fold)
-                    else:
-                        y_score = model.predict_proba(X_test_fold)[:, 1]  # Use probability of positive class
-                    y_pred = model.predict(X_test_fold)
-                    # Calculate and store the scores for each metric
-                    for metric_name, scorer in scorings.items():
-                        if metric_name in ['AUROC', 'AUPRC']:
-                            score = scorer._score_func(y_test_fold, y_score)
+                if model_name == 'DT':
+                    # Curve generation setup
+                    mean_fpr = np.linspace(0, 1, 100)
+                    tprs, aucs = [], []
+                    mean_recall = np.linspace(0, 1, 100)
+                    precisions, pr_aucs = [], []
+                    cmap = plt.get_cmap('tab10')  # Colormap
+                    # Initialize storage for scores for each fold
+                    fold_scores = {metric_name: [] for metric_name in scorings.keys()}
+                    # Manually loop through each fold in the cross-validation
+                    for fold_idx, (train_idx, test_idx) in enumerate(cv.split(X_train, y_train)):
+                        X_train_fold, X_test_fold = X_train[train_idx], X_train[test_idx]
+                        y_train_fold, y_test_fold = y_train[train_idx], y_train[test_idx]
+                        # Fit the model on the training data
+                        model.fit(X_train_fold, y_train_fold)
+                        # --------------------- SCORINGS ---------------------------
+                        # Predict on the test data
+                        # Check if the model has a decision_function method
+                        if hasattr(model, "decision_function"):
+                            # Use decision_function to get the continuous scores for each test sample
+                            y_score = model.decision_function(X_test_fold)
                         else:
-                            score = scorer._score_func(y_test_fold, y_pred)
-                        fold_scores[metric_name].append(score)
+                            # If decision_function is not available, use predict_proba to get probabilities
+                            # predict_proba returns an array with probabilities for all classes
+                            # [:, 1] extracts the probability for the positive class (class 1)
+                            y_score = model.predict_proba(X_test_fold)[:, 1] 
+                        # Get the predicted class labels for the test data
+                        y_pred = model.predict(X_test_fold)
+                        # Calculate and store the scores for each metric
+                        for metric_name, scorer in scorings.items():
+                            if metric_name in ['AUROC', 'AUPRC']:
+                                score = scorer._score_func(y_test_fold, y_score)
+                            else:
+                                score = scorer._score_func(y_test_fold, y_pred)
+                            fold_scores[metric_name].append(score)
+                        # --------------------- END SCORINGS ---------------------------
                     # --------------------- CURVES ---------------------------
-                    # Generate ROC curve for the fold
-                    roc_display = RocCurveDisplay.from_estimator(model, X_test_fold, y_test_fold,
-                                                                name=f"ROC fold {fold_idx}", alpha=0.6, lw=2,
-                                                                ax=axes[model_idx][0], color=cmap(fold_idx % 10))
-                    interp_tpr = np.interp(mean_fpr, roc_display.fpr, roc_display.tpr)
-                    interp_tpr[0] = 0.0
-                    tprs.append(interp_tpr)
-                    aucs.append(roc_display.roc_auc)
-                    # Generate Precision-Recall curve for the fold
-                    pr_display = PrecisionRecallDisplay.from_estimator(model, X_test_fold, y_test_fold,
-                                                                    name=f"PR fold {fold_idx}", alpha=0.6, lw=2,
-                                                                    ax=axes[model_idx][1], color=cmap(fold_idx % 10))
-                    interp_precision = np.interp(mean_recall, pr_display.recall[::-1], pr_display.precision[::-1])
-                    precisions.append(interp_precision)
-                    pr_aucs.append(pr_display.average_precision)
-                # Store the fold scores in the dataframe
-                for metric_name, scores in fold_scores.items():
-                    scores_df.loc[f"{model_name}_{metric_name}"] = np.around(scores, 4)
+                        # Generate ROC curve for the fold
+                        roc_display = RocCurveDisplay.from_estimator(model, X_test_fold, y_test_fold,
+                                                                    name=f"ROC fold {fold_idx}", alpha=0.6, lw=2,
+                                                                    ax=axes[model_idx][0], color=cmap(fold_idx % 10))
+                        interp_tpr = np.interp(mean_fpr, roc_display.fpr, roc_display.tpr)
+                        interp_tpr[0] = 0.0
+                        tprs.append(interp_tpr)
+                        aucs.append(roc_display.roc_auc)
+                        # Generate Precision-Recall curve for the fold
+                        pr_display = PrecisionRecallDisplay.from_estimator(model, X_test_fold, y_test_fold,
+                                                                        name=f"PR fold {fold_idx}", alpha=0.6, lw=2,
+                                                                        ax=axes[model_idx][1], color=cmap(fold_idx % 10))
+                        interp_precision = np.interp(mean_recall, pr_display.recall[::-1], pr_display.precision[::-1])
+                        precisions.append(interp_precision)
+                        pr_aucs.append(pr_display.average_precision)
                 # Plot diagonal line for random guessing in ROC curve
                 axes[model_idx][0].plot([0, 1], [0, 1], linestyle='--', lw=2, color='r', alpha=.8, label='Random guessing')
                 # Compute mean ROC curve
@@ -261,6 +263,10 @@ if __name__ == "__main__":
                 # Set Precision-Recall plot limits and title
                 axes[model_idx][1].set(xlim=[-0.05, 1.05], ylim=[-0.05, 1.05], title=f"Precision-Recall Curve - {model_name} ({group}-{method_names[j]})")
                 axes[model_idx][1].legend(loc="lower right")
+                # --------------------- END CURVES ---------------------------
+                # Store the fold scores in the dataframe
+                for metric_name, scores in fold_scores.items():
+                    scores_df.loc[f"{model_name}_{metric_name}"] = np.around(scores, 4)
             # Store the DataFrame in the dictionary with a unique key for each sheet
             sheet_name = f"{group}_{method_names[j]}"
             scores_sheets[sheet_name] = scores_df
@@ -272,4 +278,5 @@ if __name__ == "__main__":
     with pd.ExcelWriter('./output_cv_metrics/metrics.xlsx') as writer:
         for sheet_name, data in scores_sheets.items():
             data.to_excel(writer, sheet_name=sheet_name)
-    print("Successful cv metric generation for tuned models")
\ No newline at end of file
+    print("Successful cv metric generation for tuned models")
+    # --------------------------------------------------------------------------------------------------------
\ No newline at end of file
diff --git a/model_selection/output_cv_metrics.xlsx b/model_selection/output_cv_metrics.xlsx
deleted file mode 100644
index 72112a27b6ff36492eca8c7f8c5851ea62c74d66..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 31899
zcmd43cT`i^7dC7eMMYsy0Y@pL*pLzt>1`~CQWc~b2t|st&?G=0peUlEpmYgEML=q(
z2@r}%2O%OQKnO($2q7RPL<oWJqBHYOF8W*FdjI&=_XoOkbIv{YoPEyS&wieLu4--l
zZU3fCn|5r1J7P^P-KgFj1pF%>`1u|9v2ne5)x*`z6Jq7&Ch6nitP`fg`Tq3p7uPf=
z$Z*52e@gw?nKIPmcUeyMnau@6@QS}dbvQ|R;PjI+-e!|p5pS6XwdnyoO&Xi9oR?}Y
zbY%RYaqOUjiO^qdzM=bi?Q4(b9ctKW$*V2hhF%ezDpnj9D3|#xI+2A7j!U>5>nL^d
zTddQ3STM0%80)%ZwXG_0dD~UC@yerlx+j5gwrt+CiRZuLz3J*<yFTF5=heD3r*}WF
zux+mJirgG@l4`r}#It8{zkTLAHo(_@_L_C)G|i!<`|eP^)T&w_me=<6*~7yIyLk2~
zd<Y$obBQ^g+jIA%<{f9QFJ6)yuLa+mUxsyj)%JVXS58$4dYgIZ>*YSjPp5Z<${*^z
z;?SD?L^$KcZ{3?@zvQRxE+2KX+8Id*l+|w1UhHkXPV}<HC6ri+aACr^2_196PSO+I
ze`ZC?d^*~>Uta~bKOB`Z1ihAM7s&fj^1;rCFdpNngmQ5rnx>Maek}L#WZCc5Ev2L0
z`u&wpbA?&_ueiHu{c2C_&*cmetGbfeC+3A&g2?aNH*HEdv1!vjU}AinA-8YYy4e2s
zQ)+!`Xg8pqee(MQW-uRp>=#`)>t*j^i}rasAV@_o?$dOIgLzDzI{Vr@zaLnoA^Uqv
z%pO{VP8nMHH}MH(o4UsM7kS_LX-VrZ4<1I%`B?TsNZoiGWMIhn%u$|mr~gRDe0{eo
zRU1}(!(mg6JgRBm-kScP&>dPn`&+urhM)%_BHfckB(l+c2j^D@G;AzUF!5V|eY}3{
zhO5P&MHFG_Z@ewnpW3Eu-e-L7c<IBUxh3bimA(O0Tf&5*;F-Jd(}zs{3~X~vAJx6_
zaQOQB+INojYDP{C?Vm2+32H(`Z7aL7n9{+!<IEl<#YB@fg#ThKE%`}v-|e$52Zu&x
zedMt%pF-UPejB=mPkNuY;9T|Q+|HsDhmSTri+#Dz(=wkL=yfnqgH-gbgk4XedL2&Q
z<c_dM^*JAX_EgU5h1BFes?Yj(o9b}<1S{9+%>>`A^Or`b$i2M9l)pOu++^a0=>G6&
zYYa?u*(>T`uOOa$&g`qnw8A5v)%v>Mx*b0K5p(vEN8+<>T$|)-+qJLyW6oh855150
zWcI9^AX<F&;EwplyL649!*%@_7Cjc<^hO$)h*E5GiVAed+MG_qqY05Ko?k|j(<6yG
z=iA)M#dpp!MW9w8C*Vr^JQM}&&-)V=7oVpdJ42XR+C|k>wih;v$ulPqMw6a+h~UYK
zCa6U{4Y&v~<6YqqV$`Lh>dB1~WurXu#aH&6*{8N6@j;y9Re>#Uc1p#(dho5NKAUG~
z9|~8wqjB)*Lq?Wsyo&KFlPi7qzx{sc;O7L-G0%2qCP(U}_C>?;{*+h~-p#QG2G3^7
zh}U)dM8BKONl+Htx3@KlbNjgmu}6>Z%}@x?xx8;xEVXYnq2sTwQIMm{n$N=>3PsXy
zA8-94eX#lx=8#v&txw9x)on`&GP?wx9}?VRf2i=%;rkBwuR@rkVevWddBo1ETK^$?
z(4&~|F4jHczU7U*_3b{opNe=^OtWPoRV^E*TAzEhbwB6giGaxgzuOl2jHgQIfxKos
z?X|nfZ>%I(&6VmeK`nNyl7I(Y($3;x42@2H%~9^->1q>a&T^#0o9?W*Z{(Pq8FE(j
zt<afW+h$9JR4GWGU2-w?<yUiq{5G$(_sPA_`WEy~Ay5)UNS!HNP<izUCntikukHwi
z4qSpmYni!HeZ;epCvO^iv^F+e$3%+uC?vm|V%<_rnx!rG9=(r=H<b|Kd!}`_@Z8ao
zmC)v|r)qb8v4I}j*>b(ZR_^iMr1OX169#n6Xd(quda^_rIy!-Qbz4=Ou%v7;UewAd
zeF5j8HHnR?EIux;xsd<CO;@O8x01bbgdo=YrJxLR(Ht=<l6u-8_Cm?Q^Oaj&$`bn(
z<&)x`m+tD&f96`VG*Rbs@UueiZ>6X>W0&!s_IBoY^|ic|8RZzkhKF;n4^O{(5J#xF
z5aR1+5oaG=fA#Es7w7Dw;+JriCB#|r{X;gE5-&OKVXkeN(>Jt~g4S;pd-mF2`-%D&
zxA9mri;?z-5mK6MF&rc6{fEmDeTDZ9ar^AEng?w2s|^JShS0gZN=KC<UR}bfX-;GE
zZXR7e&b^m!M)9X<zMlWXpmJR7O0A)MmAW)`Z|s5ck+BAMU4xRn2rT!OGo^Ti45Y1;
zecIfh#O|Hq(1)gbKkJy&y2jB7xyU1k<yh}(%Q==?lLj|MY20GL{6cQPB6(yP9;tL$
zJHfY(XY+0KEh5nZ+xhs<j9#6Nc(#4t=GLgDi(j_{*`x{FiCgBZnA=?{%G;Ksst~9y
ze(!-U?9;Km`yc1+-gcPZV`OF~($V_~aasJX%l+$TY`XZHFFdMNk@mRIbN$$SZr^dv
zVAAL7t@wkL$1SyLknlu_6Kf<`r%DmO%BA^~7ehjYA)aVt$$RzeHIr_kr89#HZY0Hl
zt+26u;uqcoRHGh0zQdDcU0H_FTRN7YROB~%_fPanH|fnHd1$wPXV#5W8!;X=Ee?O-
zi0wC@Ko`zhgzvY`Nz2cv?~~f?d&gvc_bWU3!4UQMgVJ50*$Pj0xt{C#i=syA+;eMj
z&sl-z+h9^Ju(Bzmy1SJ7S7-L*y{`=$riE5*X8(W$THFtH-Lq-aEB{TK4*d9lc)NNy
zdET<Mz3mD4{@3@f$Wt=Sqwn&;2g3$emNr=*{r#atgW)44-rmk2DV!V#O}a0io3GPt
zcN}kDwfB)j8GWF2OwZXOC&uUCeWB0q>WH)sVM4MTnw+(!+^?ourNAsFp_WKXOY<QF
zbO?H!NP?gn&Z)X23&9aJI6d4r!6HdkPMC2>r9T;0*gLl}JT^T$HD|q6$k2=BC#iWe
zyu;B9JxtP?zfXzp@(jL}wMa!5j!jWGh~BtlZ>Dg91<b#+l)O^j%UZVWvHZ5Ua+=V9
z4{05lCoj%+b&PeO3vsQ3)>Q+p1;TSA@)&i#ortN2&NC6rhHp>TzKuD{3U!VJ%v;H|
zxl6CmiDQnQfoi%k0?k!b1Z2#jBHVX{-m+jP{Lan4YEaoV7t2^HVU{pg#<JD?=<f3q
zV+1CdzL-VwuexTgc+sm6Us~?d7l5Ll9q4Nb<sRXmsjrxEWu5ldp2oC{eO|#>H3(OR
zm&AwNn&)v;v`z~D*ufL(fB!|;mpLAUqV<b#w+@~x|NF^dU*>tN6s;4(AD*`^ta&zn
zc<4eiO{>|Jl;Sci)ZjNtz^9*^d0<F3&$L*zebDJ$Gd9A_g+~&FXf}&wGw36cS=R56
zu-`Dvf5fOCDNA@Iu|<lq8zuEFM*1vy`HNou#YpIN@sAl$f#%ILkC@f{g>@cZ%bxQ)
z2AI5E$_PdV=3w5W#?R+gN_4GU{Fo_lT1?2UOyFabmX{+x;>-p2XYeyG{Ws}w6__Po
z;i{9WZI-feDv-}^`rH}+E<$`xC1<QIan|vgDzf>Q==KFljaG7eb(FOD<C}Su-7cj^
zDLHXo{viy?5Z0vXD%Tgj70aiTYK&z@S&hC|ddws(y5zb!HoXf8OZqam8P&l@5Y9Ys
zT5P5icoF5oIFARp{##CQWQX2#i-ca={WWs46RhPb*Ysm0i);Zmta3}uoVcC@LGSC;
z8+Ld{^ykjohTm+wbe%C;kfdIn;dyI6KkEQvxW$JVJ%K=ZcmL-q&_`WX$BfdAKF21%
z@iGy#5awlsWbF<Wxn3;c^?(v>`hX&>>2f31%H3nOIAu@q*2i~DKd}x^td6F}#@!C|
zN8u!Oj~kc9Fcv4Vip063Bqv`rOxu))MDn4`27ycs-YbP4h#?<{C!_-*ww0>W2Pkzd
zU1c#(nz0fXr`ul8NEaosna@kuy2zjwY+}M!&qUrA<2YQL!kET3#=IS32kP5gr-HgR
zs$Y`)E*BCucZy-_J8e+67|CQlqUn^Pe1osUEs{u~G~+tc##k6^Z#1qdPK}S5+-h|!
zzG!c2Q_D*W%$lTXyKF;+$6^_@h@$SK@R~}b1Vwpf$50}oJjIcfyU%v0NH!lkrJ>#Z
zR^Te^3bW~4tkTX%9&)Bxh6iLpx&MvbKtlGIYal960%K?jF{X%S3FO0@15{t6WD}8R
z8l2}oHQ7LPN>w9|n5lYckDZLQ_eh|6_#PK&9;Z!QYsBmy)rv0jHGqs0x5LyO#_IS+
zoU8}=cU+NjI9_A(LLkz9m-e+c%EV3#8d)vmk;<<UzwJcmo4B))Rt4H&P14)aPU<H)
zXV&OcaC0~5=LvFvY(mbb3Un%(xT@cC?rDv<r>?IZ-*npac*E$6Ee88Vxn=2}(qP&l
zmJ`Q<?CP41h#qgXi4%{!Ij4PB^l{V)H1d`K<n@ID4~i?Yjd#P4WYEZ70nDyZGh%{R
zg!L!w2Tf-bjuZCBy}rCuWs<7c+f5yf%Q87Z`eUpaG+?JwAubgHVdI+>h#sLSX3zIW
z+MUskX_E3yd$(V`*;8ftL^tyJA;`|Rw{a30;5TwP1*0{64i+q>OxLmZE95c>bE)vF
zz?W83_?4$xa1Skme*>Or;ZhM$abJsurR=Q-HFrH_<x+9KqE>6dW8Y2onbMP!$;S@V
z!afBjkvtA<Fct&Gdf2Db0+NR)$QUjcU4Vr(!#+nJa&2jS!aj>D#+~}oXd=3IIR4yj
z?>0KftlWQ{Fh)lrb{J3H@|?b<{rHOXAIDop<CG07A0MF%r-jL4yo+ia&T+*;4@R=B
zfNzOLRoM+!hY-%~j@6#Ij-aMXX_>NXYnca)h81u!$Lrp_*d1x>p`9dpTK)N)zX23Y
zB=wQJKL(APb;N<zMn}d4hH<!%`%Q^v)($d8>hJ=_TNeuUi{qq0y;>S50loJAF5Wqy
zz&NwP2tgD50He%+LM+(0mJ*@a`x@5#z=l$QA5sM^T9VV`%}}K9p>;njeJsW9!Q=6n
zJ+EPP^=5T^*de|REgh{wk(`GJTRn37D?zUU_@Le$`l1YM@6~*Bw^?g;ZwQy=#nb^l
zn24w*hqAAza$jc!AAOpxr<IDR?gATot^>wNneAFRh-wPh*mu2>k3J`JNvp|&33{NO
zY%xZ?0`;`%JjUD&Y+Q34&p>$rgFUWQhaK4hdQBmhQ<mlxcb>L4w_-=OgN@FPfblL3
zvuD(@Gzc_%9OzRf#u!VcWgS<yI|s<Exn;fbatEpx)3+#?z%4=URV`HEq6u^h+5MJ`
zFsla5ZJ`ctwvu^u-T<Qi{RWdnb`~yAIQP|0+Ia%)vW)W%8=t}n8lMKakd47M>Ljo8
z<(MD}_$61{^fT^CQ1>|R;LD^%F!@0{B`+gTg35B*!8ceG%6R}}EXcScLEYwj*+>$r
zHEjWU&C;~h;%o+HHuJBxXezkd+>8JTD&6UWp|gb+d}GVS;%ppdR{78I=2oyD5L3L-
z*>afKh&!6Jh){69MHxO4dVG+nBCTl{Xguv{)5Y1|nB+hv5706T+%hi%DM9UW+8uL)
z7SRGSE}@&!=87|8F;X`!BkE}2A@z1v8m&7RXT|?8YSkg?wt-KuzU%Ld=v!njak@fU
z_kfK(oq%yWwo3NdJ<6^Pu9<0>h1^Es5JTkgdV`nS(trcnrPQ-}{ZEdTjoqfmeZ|P-
zv4S@jlXy3HinGBsXkxCViL3F?+|Sw2)k<yktI#Cvu@cA`mKwC*aCH-vdJc$pk&?7P
zR^OtG`*iwV+4b$a6SP_MWTW#cVejj0UJWevgSSJZj$k&%J5S0q6Q_a$zvnFjPAXNx
z`H4~PPmGp7jDaq|BuH$W+xN>zPGI`5c}Y<4So1Q1>2yA<8gFB@$2QPlxfj0}bO2i^
z(CRNbb{xDcNiLa|tchI@02Amy{e*KSgsb)<SH4UOyC%VBgpw`DuBv;WA7J5-XcZLm
z<do~+6xL#~daaU7!*VkrYOZiqcvAi5dx3p$?-18|vVdyz%3L5jNQ;Rinp1ae+Efnv
zB}hB7p6YwrPt_+~K3G2seN}SDvMJ_C`VPaLsA+QVe$Sq8Xj1Lr&P{Cx2{w#rW9uLi
zD|3z38oQ|~rZeD14BhQ(7mhZ~7>A%+SEtm%x%CLHgsLfqMPNyZ&+0eZ`I3NuLROjY
z!O?jHqhx?Pd3sthscKr5ss}lwqH0!DXc{<MQW>aNRf$~c3nA0H7Q<OJtd6mo>8Yjg
zh6eg{)r=fB$<>vBo+m>32L@5SRsO@DmnW&q@c9nAwHBh)z2T0~ngCU@8ayd!2B&9_
zZym>t&%_H`SNZpj+BRYsXjg)UaPrgeF&u4?F}rVVg0|$G&^cx`@2)m|Q)PuQK4v%n
z_4Kr39DvYf=*og#{d3<&h-7r^rUy$bY`{Rjv+%0#n!_4}wnSO#O>Z5i3v)9`G&$VT
z>dUgT>cs;X3<g1Rwa3r4!z@q&YCdX0EJ{0fO~dGFIXqW#Re)<{miZ7p@laOyRhr1y
z(he2D{6>wqiHDNHR_P+k{`b*gCSxKiw`*-xk7#RT&++6x>zO(|YQ5vzGY@9`9DeZ@
znwdWPWW}NC&_aMo7pr)A$-?e3O3hg{VA$;~gVuh|&Q%fuQO}jLewP`kCH=>r`duFH
z%zm>Ty+Z+z*dx!n=AA7|@})R+TUQuY(j&_5NdH=)*8+L*vBm&A%UH^YUUtiJ)pU}H
zjpV&D*tl7%)W~z`R;~UOi>C39HoH80M}=zllSg-Ws2?}jf5IkC`<$xZs8EcKuR31{
zuW7;!9+PLm^7;^nGpFxlC`ZcN=B!?`?6?@elKoqs^j|m;^P%}l^Jd*?tK-|hHKqLS
zxg(v6hawc8TYi*mb*0QeYi>HN<fwVTo8I#u=$CfiLY5x&cy~U*;s)0(O5{bZlKuR=
zXl@f-zOE*fMy<41Zcou9d#i|h<?yD>g34OUH4zKxF{2tj@e94L`YuU6a+{6M@_@&@
zFa8?s{~*ClteoFeX?X7Sgr3hcZi^3U&9oM3o^p20T&Lp7%1sB}m<Bo3Im6Whe(sVP
zH_aYKn%keW=ykd!9Ywgao1$X0=fa{qCHN$1ujKjIkT(#$0R!I1O65v-=%XdF%_$4v
zu6(StZ|3+Zy@kv^Gs=~uEO`}w!waiEl!-)F&KDT27Z_c;)mKqVnK7i+Q|MRapY2a!
z-Y9lx`Q{LC(cTx9vTI)q&8ZMBj&&)7OeT%&GCJB){zCA)u&T0`j%b=HJeI4o#dG?S
zj|c%VJhyn6Gr9azGp()+s-X5fKa!aor63jPzT#%Rb`8#r$=8WAdad9)GT%Z|dkiUT
zb+<~Wm>7N9@>C#iilUVxAQelnQ&G4~Y?OH}avZ1TagoHxN(uCMI=VXQ6rB9z8DZ^D
zjoXp-A4Z>h$JX`5VW63Vtxc<}=~!+Y#S4ACv5BFf8}qeEbu{&|-laxs+qC&-)Z-jL
z3O}9Wxt7d}^r^iarj5j}O%E%BDOVEl8L`vQt%E`iW7eUwX!Ak-ZT}7R@-O@5W+=$f
zCzVGv2S426OA!Ar2C{xWXJfD#KDoySh)8JiCP)7hW7i#_RHq#zb^1;HF7EI2YC>V7
z71q}oI`O+W>fTFJkAlnp4g1xbBXv}YGWgpM_X4m?Qo$J1b7@*`Z)eMxkMY$uuewq;
zM8_w(cI|p*q)UZvSScUXMybZ^#eakQ(5UhFW-BNpWW?)bLRSChlN-7gVoepn2I`?(
zPLW+&+ChD<$Ha>^v+rU39t^hVRb2lcTxKyyCaaO#39QlpwosE#lv}O~gN0SWYEls-
z(AN`a{x>eG-+;@+^^@R^)-Dx~D$+Pra1>d#5W?h*GP$$94|&K>C$<^hlFIgh1Xiy+
zBdVCLzj)BKbQatHfh`aXjar*Db+nsDN-OnC@2^SpRAKu@0z-ml+Rr0Zl<q?}0yFjI
z8XZ9w*vzevmh%c`a0j*^4KS)KM|UE%l=@F!<xKK41a+L`n&~*dhRL+zP4a|pusJ%-
zBcXl_p&MTloY(*#;BICHI%uL(-!rk*Tt(xv|Ba3Te``v6EQ>KNI&8WthV;jOLtLQY
zuaVhzM}^5^7zKw-9Lqf=Z9%WHwtO_YK)oVrPg6|GbP$s*$oM`YVQ43_XWbQ}vo;!e
zp`BShaZdTc5+%VPcOb|V=c!i&KPBGX0L=g|MJ|k^fM9gsC(vx&*G|DgBkMSoXhC?k
z7#A4OEd$1lPCRiC{;ET=#CO9pWdMlfyQfq?-Q0~1@?apv0t-db4@RU~mLfq7xn&K3
zhR8kD=YfXvpoXg&V;R0)^f8&=v=Xr+><J6NL3=wZ%8-nM5fa{gVH<{PX26)7Bs-Oh
zog8mwhne5c9zQSow9MFzDVU%)1hm|xY13U@c%qLW(6Wv#mS1nFF%zy2fsIIP_Ua2`
zhZo=HXm(O}9bRng6@VA(z3Nr13TnDEuwz@GUjSZs{lK=^kvX=!3-F=<wk<<09ZX5J
zDMf+iFHkL`vvL{9JQx`aBYCoeTVR7bJ6fRDk?&`;m-}YWj#$UGsC8@`gLWaX!+`Gt
zx=W{TU-WRj$`vC9_CHGiFH(7@j`dh`P)qNDhO`cDD+j7!lG`YjrFX%`#V?7+se7g<
z^>_MQ2)y891)C8>PZvATO#(&T*3Va)7e9Z|AvAb<tuc|_-fIi~i+M}R&QgAb0=i3N
zi<yX3;t3{=7&{b5nr4t1V!5lFyOU+{?(c?@aU|Ao<m~AXx$LS{VfFwyw(oqiFTi3T
zf6Zb@e-Dt!RQ*Fue*gipyZdI&PnSiW7i1Jnc+SAh1}@A=)b$nUNc_oBL>--$PLXk>
zE6r5#21<RUDtvKp=nUU9J4aodWuz0(t<!aZ(=$Rzutd}Vt#XC7vNFLMZ)LT1cmpg(
z$T?NG@Qf>>COtg4ExA$|4p*I$E2~mUwlG^5XUO5W?TsM}3qPh0-GkXMJ-ui(zc^dh
zA<SUNg|DFsOGyDJFIgqO)%SRk9zbJ25-ciKJmwvMAeqq;yB;LZ13_~CfU8gv9LEh2
zk`uN_tVm)>U?2?iRLBYvzqEXNVK#KOdj(5gRSnpyR<4LYpy%j`^Q}6>0C)@_N=ABm
z%S}7VM+XLysHO^&C6w6)0^Xas>A^gcHs;`2QNe7Wg^<UM?Wu+Q!m`4yt0klgQotUx
ztDK{$ZhQX}JQt=4ud4JBaHVy%Fh)3B>r?(FG6-3{E%Rz2TdUS&KO+^bVd1Acc+6Ga
z>uQKk^1v0XO~b!-@Ss%IgX2jacSURS@XH-MvGj-8M-lD^JXC-4zwEo3Fh^Z#m&I|9
zsLZ94M_%G;vZr%D5Y{{{OyaHPXE{SIRJ;%t9Q}-^YyK7};g32DfBIa><WXa`z{f`_
z^1cs?<K!WXkD*MxIDzxO9i(g;zLInvGHq@;ZJD8`;T;z?-FdqlWsfP|IZ>POaOs(H
z?EGO4vqp}`2R>qIReBz6j`dU!PTYN=EkrpaIDuQi^0?5WD@hklv>k3yxt*vg5?b>_
zT(@<HcH{Zc;%D4EQkKWYKQwMW!zIRH_INY=@YZO9BaFNVgCnl$C$?WK4GD1fUC`4e
zjNYe=s0-y*b*9_<#wAfQCVWvjx&(dIMaqIW1*P*Ltp9GgyorLPq`iiamBe<RQjVqU
zBPh`{x#rDhW{&EegEh>L?jKvezYwx>LQkhL{Og?w$}!Ehhrl~SZ;oz~vNiIlijw}6
zT;l(2wf4f9Mgx;a9A3&^Vv))m?vGkLIJA6pO$F_2yc{nDtX@Nhcg-qVEBHR(7mqj-
z1gX8=S^m59gk$CYhDDhchut!my6${bO@utwe>)EpLth|<ED%rdn>a30M+|A&wch3v
zHZF#)V({dg>yp0x`E8PQ?x<p#*@qi%37XD_(72O^4d$W6>bZ>akJ44jHUYk8=5aU1
zVjDDGC|ejS?ju@z_G;F5lbRn~n+WbH#NeVC^yLJ~zQ0gLXXcTFx$_1NxKGl_<^tbi
z=frCU!Z@W$zD8!*SJ-ig8Ej8>KWuTw@SJIP3yo3%E143#28UqeK2cB>V_O-kz1bgX
zr$%3YiS-UNPQGv7K)Ytn-^3Fc+-}1k8QpGEYvXWMq0z?WY;VIGkCx4mUF-kldsg9T
z)AB9k!MwlXqnlNfdP^<V<`xrGRbF3hvz@4j5h;VIO6y(vVC)evdaVrB7JeY#q-o+g
z{uwc6!gM#`KuBPLt@foBG4uF@=jB|Z6Fje&t(t<##msL?9FXu2iurZltckdt=l{*^
zzu0SFl+jikHED`B<SzY06(|JAjDH!2Gsn`h^%I>ZY6=mDg@7+@tblmLKP*RMR9u0k
zS6(Wyt*;KTgAe8qNB@a6sjZK7dq)gRDlSLf^yokQ$#YRD+vrIfEqX#Re*9VXz>2+7
z8}Gj?h>fPUX{sVoA7W1nP{`f2U<LtVYWAkC>e1U3T;&?2jEm2;K~5VVXE&Q^2`Au~
zLq38ZJ)@xll*gbKqP8=1N$PO3y;liI1ce)*zHe#G#wUxzdY+H>SH%6Bg-@%EGR#bo
zAUh3A!+<Oaston*;vi#uxsJVWL3oD6Jc5PfitVE%00}Af*(y)H-%2_*r2x~6MT2c+
zsgGJoI#UYfUrIgUH~*z>ju==spX!>3k(G7D<&WRam})DHlKPi&(0(MEb-e+q*;{Vj
zoHWvPXeI&u2m3SWda2QN-Noo$tJaqfNbV{(*_WOR*IAfrPT(5O6wDb3s|UZ##4U4t
zil8JXvZVhLXfA-L*`W)er6hjf?KW?-E|7Iz!Z3&Voz4g&9b_MoS7mCB7Y)0(yr*a=
z3<#Rprzkfy&5IT<JNj6yzK!=O$R~RB+shE@ht^}hM9?H~K>JLIaQh2+SN)WxZ9&sl
z9Ruqw{=sI1X@MsQk`|?py%k$<*tB9wvLqbT@A!@7p4a4tdc(U(V#ZP%{EQb`1cc04
zeed~z)Fr#KcuGh(x5KSQy#69YllZ}}sZq5=gj<Y6+<8n30zy9aNhnJ7k@z0+aRM3T
ze@MI^p3ojMU3|+M6No%j%$W-|R>w>h-!jHXnK%}6N?L4aIahpZJ)Nu>vtc*|sVIp)
znv*skk`@Fu(q|H57mJ;77%9_JSS?A*4c%_RI`ijA-Kgl51R^tb*W`d6>?t6>TXfp0
zVv-APSsk+r>uj8d%5&s^dT@Xp6MH+)7JrWh*@n#Pe+Qb{dn(AmD8yFMI-4O0uq5mc
zwTGVSnY+EM28kc#19^0!i-T1Qf|I!_SVTGo<e|aG#x^Kctmm0n^iPa7T0e~9+ko^o
zyI(FpP1D>;C4HweTnMaCcB^8~r>bSd+QFbY#@EkVzyop?wM60X>}i^;!;Y_l4wFQu
z2@4cH@M+c-72{m+a+by{cj2f+!1pSV0Gq+?C8aToT{sc|H$Hk&#=yN&+tGShG_mC9
zOMQ~(tTp?A5Qg24Sp|{?x+G7=2Gea!CAlJCna**tSj89unmtP``XuQ)nc+g9faYLX
z#a3(SMsW-AfW%S!=SA>{oE<E(5Z~d(|1hRn<2Ov7po^$w7Gm?Elwj`{W7|P<E9mS1
z7`KuzdluHIG<GN90GuS#yn3-ZQ_#6i5VWAcl2segxUg0Gzf^l{1$*Y=7t{){*j1-r
zBJiJgfgYOZl3|(E!MmGEdkp0G*|#+|H8`hFVk%^2@O*_iz~8aY{A$bI)O$H3LFe6o
zwj_STx1%Waeh%rF^OKkefF@@@rxTq;t*D5dha!WK_%W^x#tBO*g5ywRDqsZf2SH#z
zP!XJmQo5+L_zmOD1I9~|aITX5c#pE7rStceB#ta>o(w>eu<s;GElwy7+Ow+vlB0!p
z!{nx@MRn&|GJ`jFkz_V(LT6x~C9jh~MDU5aEJm77Q8QIp)tMHC3#pM+l`Bd}{sP#N
zZ*c#NsaZ^wJBxS%Pm4u;t#a|k(^#uD1T>l(-2lKfVH}D49j>X#1<=*l;hM%-Rkb?+
zTub^jTsylC*P_W(kA2_a+AIIfCh6^$p?~p~5GwGmdH}RW*|s^iSi;&Vb{(zNInuKj
zgPPxhXcmNOJ*P{o9U@iVfPjHDJ8xI$w>X0*F&kp}lfChfL35aD<yv_uKES_{RZuEC
zI<h!Cjg<3W{yH%?+fp;OToaIjC$;q26qJ|aD@&_<=+<k+IBpEgd98;$(C<2+s_eHi
zGCjRKtmZ0<BVizrG=4n`AMg2zdH7mzi!r3qEU<5-sFyXjt~y(o%S+~#^F}ip&@@(I
zNGmW1O!^?D3t80^(1?*kW5P9r)ecpm2eG8V2EETAvHX#<Z)9<32HstmSYr_g&y~gD
zMsgn*ED>=Pu&9(Us8HZaksprRQod?1S6I%NM5-F}??DCl=9l>ZSaNnboJ7Xch<Qtk
zI$o6ht@P>TuFwz~(Vlz6n>ks>oV_PE(#{zCt~(<bcksssa&ZK#xzu(|JTz5#pAZ&D
z6q%ZMsIKxpG3?7E&+6^k0wEq<jRpEcA4O}!@XP5U-RpQwR$+Ikz!udv%6^(l^r>5D
zG;@|`sbkveBihQy8w23A+yB99f`T@4{>3*!wR3Cl6es*qdM{%~ZL^?u8NFRtN+bPV
z-Ttnv-I0&vlm8^xA4xl%Bye6?H=_Su&pEyUh87Aw7E<~m8I8uxOhwBo&3hI~#}<kR
zIgi<JTi=QHi+dc}k+}Pr!4Zyyxz`E`SDyPsolw}v?;WvK(Vc6_=H2z=$T}-szVWf&
z(rnz}ntW#<zI6{bv&?iK_19d~Z7TKr%IO~67^42WpUGqD>1G2HdgPo&f^I~5VN{_=
z)ksPxHa>Csv=X_r*B6!ArfV$h;NjmlSB|M2nz*j)rla!G!h7zDh7g=ApTG><?kDBb
z-NZ~##swevoiaQsa6-ZIh|p0}0qYVRwF9fHO`7X3V(l>QDo~6nKWDZRwoC54lGJmf
zvc~4b-T}^7y`Xa?JI-_~N6R%Hm5ud&BEE7m$x&#zMtNa;Y<OI_*z`-dvFbi44&G+5
z)bmnG*z@i?OS@wL$|K+Q?bQSq<@#X%2Wppmb;frM14@X?NH0Wh>hTIi7%9iLGdg$l
zUWL)pRg9r{y_qM}!U2NaL&u)^dNZ*}-8i$#B7fXWOu1}bie@hzm}OfDx2uFx?83Yd
z1+_W2Xs5n0cT}p@3tzAP7e273UDyi$C}Kl2p^2U}jvVV~I=nDk<H^`BLwCn}93*SM
z@jT`q@M*;{mWmR?@h_E0H!$8S&0Z96w8*Zb9eGUCo0ggTHir;l3|4Dvw`gk<B*OxH
z8FF+71hQZB1OZQZ9oupE(!yL*bHemZiS&!p7jp0GZvTK(5VNkmp;}hcq{=rp`AlT&
zgz-ax{AYSoOEwp|onAUPeC{YTT%ki;2Pfu5@ag3{Et+^*6<yk<2}w^(lC`@_V|T1H
zDI|&;=`jv-HZ(oVn7VU9gwG>)^v;sLxg~Ml=*Du9dB>oF&rVxKe&`d(Lljy9gj3A4
zI+Czii_gR=R;0|`3NEup08KG5$}YrFAis<qv3$-RnckC7LG(!$T*&*nTywjr`Uv>T
z_;{j-ye>iG4RRnffvhK>kG~X=FNWCA+tke6b3sonjVVxY!L_;@z8o(bGXp(#tu%9?
z$rP^_R+@?OkgYcRKN6?!;L1~i=GF}l8A1I;Foe5DiGh%2uEttrAKVLwm~Swg>)S!Y
z(ttiDPC3uMf3FOrblJa-f-V!VLIeZp$(rbBWt~{=-LxT=8je|(aUx>U-7?5+CHFy*
z0syaivd!NoeNMey`Uqr8hsO70lWUUz5t`D7EcnX{ceq)IIOL}p+dTaQ3_HBaQ=?k@
zQl~^=6|-$6FF^gS<=t?wn@311W&hFXHf}}$#lJNYbmxIw`w+LTFxl~+p|*L1hEhM@
zf?H`SsN)d7g$voV7=+2+TnBpy(uTWUytHQ2mA8vfngeP%4(;nJCwIfD?(!**$Ae$h
zDZ^nNp>xQiE*U=l7;JRTxHbQ4*w$N4ZhK()YO$>}7W7CCC=xKXJh04|-H3YET%+}y
z`DpsGN3&wd+ZY?6nJ~17*;8@~h=5Ol2jrAt9g5&R<Y@@)3EB|#<faBo^hr3EbAKrX
zR9~P8qJOE{iY2Yb&er&$&l@~V6c$?e57WUO`xT}7N%WDa!bzI6AV3GiZp=$XO7#8E
z0R;g%Aa-L>s=q`Z#c4h8`&r-|`gh<w1U9m&+O?RI>!>c~u}i}?a1_?3|E_F3a_&OA
zG=KpO5dQ+kUq{X+L?C>#?D{T3W1k={0Ji0ls~2%wRMtgk>}L`nLPKlf@6!0Ucfyr6
zz`@>_UEPToAn;RyTY(z(#_)9)LwguAFTab>*iSQ03Epg{|06<U4@3dfM(z7YgvNd@
zu<G?w?@IhYZQn&`?9-l?8Yn^fE<!6@7oo8m3sUb$kk&<LknbWi_SY<pSuJ*EVAc=8
zW;zH60cowO7}~PTSSA*83<ri#W?IY}fOq)_{{pt@{0-aiqd#-KmhY|?f#2vrei9@y
zHQgfcjy`^L6700j8L$=v{ouV+BmC$b$jE{hr25lofCQ4WX%UcQvd?^^1TD*TJumR3
zSY^Yz*Z|H)`r&+ppE=+1ci6^={V8li0bWAfj?4p;itLZP2Q0ShkDS3eNd*3KDVhvU
zTQZ?{g~a7HK$7r*>M9@=!MW8Dr}8t@K_wWU&i=0HVs{~DElw)wZ`4NH@XWQON8hoc
z>~<|WqVCwnUH0CH_KyUOeXHnVqw);#_W=dzPr#rqu#XJ!_ntQC14d2|QVhcu0mk5`
z?UTSGCBTpD`2;+Y4{LD<z^>qyTt8aU?g1y?Mtr)C_kfcxPz=C(U|cIW)sE2twi)aF
z9YBZ&kaz4G0-G9?)3!c;V&Kioil$N8F~vfGt_*G>cQSlhIGGGr<$`J_t&7lB=ze)?
z%QiSx0rx772FF#90>AlsvaIG$a~)V2LEx4A&j^iu6>G9!@w1~gH?k894xl!zU!yje
ze@Qltbrl*wvTc5PRwUonz_w#tw?J|{kLeQHQ=qf$8Gn=yW7_+Y)OXm{=fCqi$=3Q-
z_}ym;v2|odFPTWCg|77vswh_|GRw&U>!m^+OahgOPM?vr=;luG?^hlebZ*cSR!dYv
z`I-S3$mMpR_kg+C?te%&?;j-Fe8VC@va$Y4vQ?7&|3k9j@c)o(tPUizCD#4kY{xun
zon*rZC&A?al8yKe$tLNIo_YX~Y^%$+zs`nEw69=$S5#5oNwzp$M-(ogDsj<WdUb|C
zr!P;#`?`Eit*v5-g&v&_Ej8iE-ZO~JPsSOf){bSQCnKE1O(J00jO-z**=hqT?l=@~
z>0tMu&~*!dWOH_1C)p}1eG-J$Nwy9S*R+%wA(p(q=JXGet>HV#w(N7!IQ(>j$QXTx
zpR)C}aJL4LW#oM|A*oK0_p=XeRgMUSn{<hMntj-UylAu+-@~mSymcVGUnPB|OsQvV
z(F%EKwFo6ZM<jI7qL-E|;Myo9AkS_U_*RZweNM6xDxv5uNwj`vW@9(XHFo58G4DUQ
zhIC8R+_#)ddKc`2P*-@#?{`^a>!?4#IY}<`#?9E@3~-bks4OJ*`G2rY_-yJgvHh~#
zdnTv5sA=iFbGi?ew@JN%EeFd7{~7)#*Y6b3MKYy2(ApzR*Z9L8;q3VMx&TMM^tA1E
zc84+h6{I<R(>_1CHF_ttJf!KuhgzY8{(Ekaa{PWqlH25tYeMH>oL6>#p`?6xs&ZZW
z(~SKaH5m~Mrt6Q1ED;HQ{<(o;+iz?6<~OSBz`Y-z>m47xa}Fjxw_lP{cZBh=g@3<q
zn3(&y6AFJRXB8-)mm{9A>CtO+b37k#-^u1JSJxhC1rh6a+DX|N93LPypN*zo9l9u>
zySpTzNt9<RK;lFED<g;x<r`9{@=Y3|_7}fxf$7FDcbX=}KT8hmOuOk@axemybs<U=
zePOC@1lJYwnQJ8TW;1bam-?w!?>V3QdY!kvWI%m2>WGt8ndnlT=xVZ!ieH#?%RwwO
zU!5?^brFju7(rddtla^iHocrZ`aZ)*Dm;InuzbAJ{#f!LJt-DB7R&qq(X(4MYFR0u
zf60YFOJ5+N&(S%C1M^gs(9*@?<_~DUuvSmq=(bvSX&0!iD!wN~HNa;w>2;c_i}7Up
zpGmUBAPUVq&%c6KJa?OMp;^37wFj+Zz<0lR@~jM`u1y+lKCULVM9|Ge0wB$sZiqni
z?1oexA`&!$Z#?<3oN`{`k~rnmKr@eL?^J?-R3pD4L_?sf#qD;KmhZJjo-C_p1|p37
zFO;dq`8yM$UPwd%pq-%5lu*d9gA;$IYw&6MGR|l)pS3vq6@|mv+OA2f#V4k~{BAcj
z^2{N_J!-EFYHKj%Z@d%|-Ud5kmugAq+rvz}H{<;QP^3-ZjR)<vfs(((Phs{nr1F^N
zU}>DhPbjdK#A2uEIPq;H&pZ&8s*b@bIuX01OunsFI}dHda5<jcxrsOP7EpJ3BO=wo
z-<F~erg*!srpe#{>}N0#^OVr>=+2Gk$oz{ykMIOFUG3Eu4K>G$>y~EP=HuUOL}|zJ
zoMPWisJ<LOg_CdGz}nptet;#BijKbYAbSZ&fG`}uYj~Xq<>wn&m)5X<K!Etzi#AOk
z8|z3Xh3GBZ^Zyx0>nf%aJxhS#Q4$;>&c^KRF4pStERT|u14Dt^TKlMdKbSQ^+(zRe
zYM%>0yPd)d0);N@pMD=LQ7VAZ6ab8d00e35p-3Ph`)r{v{soP=+p)wm*B%tJ4#yy6
zU5SOL5=RZRrv~T?&V{>O<cm+Cx&EA4`^M@(XptZQvqmV*1O++c(9FJ-wJb-I;PL)$
z(2(XOTwDg=PME>J03n+W=$UQBoxn5Go|b-wq<N}?jJnwCfblgDZ{+p2fqK0sL318}
z+usDJwekLz4aN%Rfoa>Dd7VDvX3ZeuTJRN7+fan)p~$Wuay0fe&rJ=N=tDW*;02Ht
zFK9q*#n<!I@sD*%GbNm>H^ex^Pw)9?e6nPY1v5as#2y8XBfsi>17OdMztw2$J7b;S
z__t69v{(B+Rqz2msHgY6PzYq~(Fcr^GTVL>3W1D$`jvb@l&t=}P>B73y4WI&Iv%C-
zccBowag{ililWmNWsU)bLgQd^Xr0{v$RVBf$yV(6&)ALiFuD7GgvsUrwf4&}8UOPz
znR{SQD)}Aa*I}~W{|u9v-XGz98?ap|fbDQ%J#hfDmb4*uEJzKa({{-o*V@Rn1Tf!z
zdiMaa_0YXwM<us45E|#gE?{1ftLiOh6H80(fZX5&bXU(uNz0oB$b7+)K#&n&b%`IR
zE!I!K^*SN2TL$a_AQ}aLXm`P6s8?<(Y7ekp-#yP4+$-{HOayIAMq=mNPA;=@&}K-H
zsB@AxEzj%V`l+q}Aa#=iSerFz@4CR45pcZ+2zq=_s*#f05@-$rvU>C@W7d6-vq^Hp
zUN`~mBNmV=(`;T5|A#T7N3s+=eHy^@6_~$EyEWXJ@+(>w@xN%@b>I6ntqcD-t?T2r
z3Zxw*{%>jBVJ7aXYb;RD!+uh)J-j#pFo7KCAb<n=3}(-OU1pVq*ol4(C?UbQzY{?2
zFVp-zFn<HNC1ALD0;<@PNv=+_$pDPOAYagc-T;j6({9m$NB)dF%R+FHa@Ltia7&IK
zEfLkg$@fcihvx;*-8Rsf$+bA-`np)G2M*K;A?h~pZY2F`qwhLMUNBz==pa{yT(9Z{
z+Au@V4SG0kJ;-#oaD#+oCA!jZ@f%S%XO$XB8*7pLHYiJ?jRi1<$LXyh^pUfc3*qQi
z#9GtsIMyI{)iOv3iTJbKfSYsErsiK)3dw%QXh5aV4<RH_DWvYX_Y+@Qa~dJh*=*{t
zu)z&zediXB`-sGTlwIDnEDL&7G^=39f-o`B{#N*j7u~CnHL^61tWaVSMru0d(eng6
zIaU}VnWW|qQ_D>X?D02MdphS(h%%$qyW+h02^LstS3~SmAz{hBUgv<>lCmF#LWK+x
zHLEolAHqzhGikl^$>HHyw7|g}VOfj81i5icpqpPH!<R9~!A~@{F!H7*6Y&;RPYyj9
zgiSH#XII@>3;tEUilkh%3WY$@%7O7%2)cat5-os5UmBtoPB;b(I<Ll6&I%7E5Yi%V
zsAa27rJ=Ix*I31srfSOC_=9?s!9fNxDNNDD_Zyu$-Kw=|WQkRYgbg|i0hL1TjHTA4
z&nr_S%N>)15Hf8n3mRDU_Hg3s8#E)LF{zbUNV{E3G!{;S4CrVCmM*)0i|V%^wszH7
zjNDc87k2kle%-3^c7(@V^}KL6aBIVuPwlhtJOHdM?eJFqd`5UUTtj@0XV|Cqj4%)9
zqt?X>lJ$2OqBx9bl?ykvG%$X;G1;4RZKlDLY@P+&<xt=0Ju=oIJjG4VRgrI;3Q->n
zrItS58gs<x7)OzZ-RPZH=+w2^9Lw6FT%Pfpl!37$%HFr--66idJER)U`H&uQKK9@U
ztyESZ2Y8ejM9!}o4MwG<NAion)6glK=b}5hFh*Jyk2hZudd07iXZ`Lp|M-_tq0vo{
z?M0kR5zkL3JR}_#N*pavKE?{rTsAP+BK;{;r0Fl$+B>iIdFMs=nLONlWBL`(qeD_J
zO`G{29r$Se7}}Fqwkz=LU*%KxltV75){ZI^>G*0sU`6dtPaCD_xNAO`x*S^}-*srX
zM*GdH+B?(w{ed@x&P~1B;W|#68&Z=}NY`y>N@>EATWZXkn|QXy8<|~=GFD!wv-gm_
zHqW})s;ASb_x1A&`w$n1(vtkXtz&b!`qqDmq@NGn9D*a`O80K%yXQYUy0r{0I<5h$
z(^n<TZEpSirA}n=BcxXIs1eum)+S1TuB2XT6GScAZgkGA`lzt&AI%Kb1!bgh)71~J
znh;_Ig&`-?>usUU(Z&dEDe7U8zOtUkl&lsqUk~Z-k!q-obT3J8suS%WJyqCNw2w6D
z4}F2imFDqIdwRK*LTYZf8elZ1qBLlvZS+RP{sn~bVjZ(pMlBV@OzNaPEeY-_h(WjE
zc!}^c(cvfY*{!izNrf24#&EI0=h7-BU3M?z3Q=E1V$+jE>pIhE`u8F@7tje1%G~9z
z*9a}9%JMpG_sYq28IM-XSRX2eQmkY16BbYE?lF3+6FKLqEBP>>f~7J06ssT}UMT$F
zoa7nTuuF~V;SnNbcF#NkqOBn3MM_E|$K7ZVKF%eCn8pXnwQ_-Wfr?=v^A_z&4uQmp
zypy@Zb4{8oyn^Yw!PVtetg@*FFY;NWSEh}a!t2Nx;_|KgW-{?9DIda)YdtwIYS1A_
zyPz8!E7Di@gui9@OLjQSaPAI=w}Yy2MwlF3HNa^o<e1J0Y+S$s9rfTDxBy4d%K*}K
zGk3u=EN$F|?OudX;68>-cY~Swn+5x)-c^IDZ-63%-n0SR2Fu|PuGR$hdmezI2<OQZ
zor-5+)t|U`#%t;Ff$EqFG6H4kTNIB&bC+9#JlXDu00j6>lW{r~Ho93s9=TfVw?O1y
zUb4_mJ=ojrG#cfUe+eu7AsW<kZiWz+j+9s7XfAD*;PL=>O|2e$(~a&2@LL}l3*a~F
zPyfQ2AEgM6K&7*Ilf;GshxJl~Zafj+MJBU41X+BWh~N^4;k4zhZYt3Uz;93hzp<B4
zjp!W9&WX(<`iPpttwFX99i8kPxqRa7SL4|JdQe$VI=asU>UtA1q2KEQC?o%S+W?>F
zKQ}#$NXQMu0Ln=Aw;9ybYR%4)D`dsLIsk8Azv%%eRw&6x@}eWL+_;{0MbJ$T-wTRz
zvb>8}H(v=z;=r{%!$K=q$BI-M|L)EpCD2U|>do^y|CL6v4+*&Gfja3zqbc3079@hZ
zElrE>?sDpj6I(r~^s{=R?k=%DAd}ES0~bK#Wh8ph5whDki*X|`a79*6XG<9pa_C-a
zJ;|N@rVMfY)fs(#Q7SdnSsQDH@SDB~S{P3qR5pf@?_l_gaw00#(7Cy|=RFe039~P^
zr;bE62A4OX0JH=`Z1zbo<6a%Ms6gpxZ8?Qi;Rm}>d)nu&b$PCXAE?@1y*WjPoQ3hm
zRFTi8(@K9<O;kBhUT>Dsv1nES@*-d(8(Uytz(y7~20ySmZL*7)MLZyWyDte19&2Dg
zQatghD-6KD4jdZ>6sbdYdh3mm>h(w$Eb#+5Cqc#w^J<_X&G4Lp@rIV3KU#_str-V@
z*>e0Z1!GC}D+ku^YWX%Ag`AC#`3)G+25XAWe?SQPQKx|}BVL_>sjV$~pDf<`vuTj&
z%p&d=FYdbwaFpyzcFSPt;3UQA@pq2$zjwHpyjhV>)n~>NNyI4eg+2uwcy^o87G*F%
zoMF21PEwNn){6B7-R2B~SQ->j?34fi%-*pwa96<?P)$@IdvaZ520G{RGJ;zFt|mfz
zvR{?){q(h*M{@Z2?vJ_$x|{;2jOw|k2i*3M7rTC0fdar3)xG&wsQ}d9^``BTRa?I+
zqYa4M*^ZkDD3+f?_^v9do;het!ZXvy3q$DX?x7Q&FaiNPs4C4O$>CdLfoh_`DFnJ9
z7NUkR7ABG$-hD}yMUz+26Q`Nsm<`oL3(g0ny#b=*&aa6MxeY|exsX4uI9p{pFZW;R
zrU)RaTzH`M*Eji|d*nS6V81`bs0z83vtDkW)X@?DO4vEH1%pl>Ojg4XNyclwC4v3{
zs6}5qYYsO*KhB*@<d!20x>RNhb1W74B1?L?l2yq`YJt8t2L1b87Tf|=(b|0DfUYdu
z-Z&PIsqiZ+>tD&QsPZ$b@<Y;B8ft1<I|2uDXXGp_l645M%CD=_i$L9e@O)MaL=Hzy
z?j~7O?mo0TIcWxnqo=J{-~4gD{`8{Nh2^Q%1y^6Me3FgaC=pL5j^7KncxpUPvYTgq
zyR!0a+4)RT=W+ncN@;q;?%QB8l}cx^O0tHpud-Y)_LR3HF(JZLQv<Xu2ij(E=y5uI
zY*pA>7?*63oYMwFnFW?DjRY_Q=4G<0%>SV|B$H(=EL5Hv`*|%{tqjq87RX&cO77!f
zNoxuM-)e9?#0C*N`VLQJpr8l{pUaT<m4w=fA}h0iOznt3xJkFj=h=sbDn|svOWtbO
z9LooUbxXg0Ri*R-x1!l`kor}^FpbGry?X<D3d<_asI9f^$#4aPcn~0xMj+gtHF;+(
zX(zN43!LVCen84nmdDe+?oKX$$IR&A+D_zfN2>gJ`LZ4PAwIXmth*xwPVa*}72m$l
zp_}|{Dd7BA`teb+pI?qnlbKm=dAt+QKGr)_Aa8G^W8EFNLUPpF=)wovvN`N?KPtss
zgd@B5j?d)%NP&-`J3q|*&WS>nu@-Kpr*YV&?}Oaeed8{%{q{a7)TP~Vx^=}Sr5wI`
zWx4~WkLX!&o4~Y3pEP=!H2(2?tKD4K^R1G#p{HA3#w<TsVE!?o*V>47^16gv`utq_
z2#49j69vz=j_%8jg}5!{q{i+$GIl}eFRmRP9CLh+HPTORf0J@Sdnu;P+D&bO-efv;
z!rtA{Bj7&M_L{Jx$HG}?g3sX>!TzfszBOkb=$D{WcoxYmo!z%}8CHHHNZ`-pO|zYc
zBM$sIlOqn<Qb`nN6c#GWD5<TzZEc-%OBE5mK#XoXVQqAZcE3;4PbwvMz)1TQ2FE+k
z@5t-95ACN*=Y6Dfl1PGmd4X93%IYOcIzJePWGTM*;Fap>PD|5YGOso|W#1u#qtmfR
za96YB$9wucGclL@S4s!?_)7T!#sjFu*KJvO?X^mbZf=OoxIQq69WcsGaOA}pT1=d`
zg9#ho$~N>@(&W!EzLwi@=R<lV>Pt;XdXl8$<Osq)3r6gJM9j)CRXv~6b3>J~twlq-
zb>Gr`{)QOt7mFFG@sxYgp(aBL(upxE=_mLvRh6&dms`D&J%_!WQgqvb1l)_{6%QUZ
z*6?^YJ9jxu{)Vui;02?2x1!J#-cPf>&Bs2?Uc2FRRN)~1l@9&)=_jDy&=vy~#Pm#f
zTubl@LYzZ9JpNmTxrxHi`CR4x_H!{BxG({CWkc&)%>mNvw~ON*nWGQuG<CzvW47t&
zWsG(pP6W;jg&gfD?Xg~YPQ;m>V^3(d72lq4z-t4gL<Ge&!Jm{8)pj66*Gq}~xcERh
z)SgfF>9q4ozj3O$#7P46`Ru89OD*0+EdI;y;;c88WurDAX<4L1uxB4U;X(njB{^3f
zTrFg#m63pjQr5LV6+>V$WbLoag?dxG9H2_8aQ~Sq4dvE7eQ(6`yDIJ322~nIU{If=
zE1*h4F8)H5wqAk%YgL+`r=K6gM=Nldut6%)uU}<HTc+=oEGGRC3zjqZXzfG;gooBI
z2oH2Ix<{+*qo+L(FtaDaIwV}2yWuw>K){>`1bpnv7lw-}FvUkz)Mm*I{2?a|qC=`)
z7{vpEtPhJ$+hzI{h+wmumI@k2?Nv1^PH8Po0p-I&po)NetbH+jo633xKD%ND2$-J&
zwM75-jEC#M`Cl_0drCiB*#3W|O7n}-Z!QBEkBF`ocmoQo%^-eiC_@MzGTw&FHLC+H
zjHiSlpkvAhO4@?Gz|!|jpqMDz370pk0BDM6;QTQ0Uf|cii-}ws>cG}M{qnN)@_UyC
z@&?nViZTRaIXzvU<W2`AmbtLkF}day^_Cf32s5zqCq;si<@D=XA}=6A4%|$I9oe8s
zb0z%0kp%&oG@y8V!};)T8AuL)&zb(Dn!{u;*Al<3NduxicBfv~q*aou5#c}Aq?LQG
zYtr}tP1;7@VqKH=y%@m{ENkgn*Q5a@4vd*s<i!m#rFBgjAX8eu7;U50XkC*=`h_NK
zy8)m{>-%vrS`ZV=Oss3t`o3$@@ITk2_5IMKaZYb62wE+^^*=Oee1B`wwgCY;`@y@e
zNh=}|zX|+QlU9w^{YR7bb2;R{7e842eLLoly3xPokUunOLqF4`h5oHc8`_{rTg?X)
zkt9G7$roY1L7}-$x7p%<+{^cCqkg`*$7Yb50L&mzaRyvn@LoyNb3@S&_AFrhHQVOg
z`@{RNqu}}*LFWm}76b=LWfzd^-%uat_ce({+m3}ExQaL5kk4<!UeBYf-|U5G+Q5L4
zKE02j@h#Q>n#cjkjjA-&|DUSTp#Mvi7WBVVX^fLURi#y9Gyqi^`PZs6-CwEFmZ~AS
zLV?l@ZaF=icQ^qvO-z0uaeVNp+9|(fs+{>6y$LaI%fI4^rXl9@m|y3Y$3~VN1WtcT
zZ?(g(*(iTljSaJ4akHz^hA&WK)~|jD`ZZXSX9sHl<MC2K^NlkbR>Nj&nhV-%4`4ML
z0IN*~I?W$@H<NfT=~9Yso&?OlS9x`I!J%RxiG>Ped|jAZ=|C`-n4zKR-pSnzIlVy_
zRgCcQr99vAs$Sh>)qrxJ<&~03-zw6V)#=yf;`O0SG$~|0iyy#hzU9dLB~L$ZW?sO&
z&*H*3i`6<3Xn_G#ZLlOc;M!{+Kjv%)fYm~?_>)~#a|2<nfh=X0(<o&AAc09@(rY?s
zW5ZMKtYOMh))=#-tly>J&=L*7qD%~p6fU)|RQ9e&1%4i^QnvVTqPP-TMQUS~yjW!r
zmysxMxfw^r=rw;5)mL${oH7jD`yhwkls3;~jX9tyN}1y{CM`X7JZrv?LFJYU9B>(g
zSC-Im95QbVuFH|oS^VK>h<*5cN(FKjzB)wrUm&L>*N=~`mN~-AR%8Ufk#UdypU%!Z
zp31k6;|IxJ5wcaXB{P|c6poUNj855QJIEeI$d*x7*<>A(m18R+dmKa=Cvr$;+0X6t
zd-6LsJ%2p@@N!;W*XO>^b$#z~eaHKYnGBYAas555kz8s7_l9Q!%Qf<Kf7ukqWs(On
z#qHc%i%onlh~Lx3uPX_hE<;5as^cwa*}l;7?q{=xZwmXU>{NRtt+?m|`Dh#H&*rLL
z(l&BFKt9^weEmiG&jrJ=Evx$EV#pxvV85DsLR34nd8Z%0)gWhdlSAs^@e}D+DlK{H
zF9ub!%ruvqL=rsGn{-{_Nm!b7+U)<PQP7cCzEM>wmwf&vKmObArfkLvt$wkFv~EGS
zWduV%J^Dq7a3l(IU8~aY4r$>IV`QA-e8^(i^h(8zvh?(}_tgz8HGKY_5U~r3TQ&3q
zWDc)Xe{no`>9MZY^Z*f+EiCYh*TQ4{NL*fT8+Rz`QT(*%2kuao$Z6^bHk$kn1T1;H
zZ??KhsIViF_1;j~MEC6wKV_{}jo@)x-gs5YpnzkZPK%}4N0S(5&_5r>6D)mWZK~@E
z>JJeqc_eEZBKn=dLgvkR`FglTJ7vGnSu|XQoCS#@UeI5>vyyx2nU)74d+p1TpREIL
zVf=0KP)FsAHb0(gcL}e>qrH{mvUR57U#Mb=WvY#iF`zNz%Br)}=#YxS`GLN&(l>&m
zoRkx8^HdIWjQ1JGpWW)o^Xkca`tw$EhHjuFe97bc1>MC8?Lv{c*%YX=fx+m286V8S
zD{DzvXo}D7<<evsx*dpWz*5?`puEd%!zQ&IH*0uawGFRh)Xwh~J)h<iVwdRE7(<#5
zxA3dgOb4{ceAY47m`*Yq6Ie^V#-37@MW^+mT<uy}z#Q!fs`!a>Ii47uBQN2$O#_!%
z3~Gj-n6Ssb^_*7uC}iOlMk5f;+MHW&W!;;xmH4z=p|%EyZkUxx<%fqXJWh=&dhMuq
zv|viXr>`7+?(!|>@bKbjqwg~uOHa-c#Vt+DH#b#8dI_^YomA9Lv40S5n9ZmfO`01i
z^7dLTtG-auaC^<sdCK2&RhEfSrBbFOUC6XoZdheRpvPyaaK_TzjdxI$O>>OGLyGgQ
z*Lfz!3v!vRNv8x2X2+q(pjz=CCec~62JP}I8)bSRWEmW4R+U@+jaPRAPo)=r5NM`c
zkt;dQdyCWsNIb)J6A~q=T5H)AYaxD|dTc&*@V`|6p*md)u`=mr-iphwrsusT;oO;Z
zTQ1wQO}zFNP&@>73U8O>GPJK3zD&Y!O+Uf)_H9LE+z7D79Cz$^qsKWyhzqa=oOXjZ
zDMsAIlZpt?hG=>sFlMcD^-Je<Or~uO*UJxok88}e{WY$ENsM5(W0&iBu?yb@0i^))
z*dD=JHc;&V!5TQPv6~IU?j#=J@jd7>U<c-CU(>KVt%nAKyKS2d*nwG<8}?>ZVl7E1
zb~Fws9>Vpvx@z%SVSvQ6F~>=p%ICOlNZwtHW$tt5Rr9i13mz=;;oqp_-=s7nWXgP_
zpi61EG#FZDucd;SWwr6S#S_>j7zT2xRuK6>-;DLFN&dfqjee{D5!i^~-`RF9#ARyl
z!S#+8<mr|Iws<Hg4Kq7$O-g^??mZGbfr#9A=hfR1glb{slQ%$9zr~)BabsXmZ79aN
za&GO@dr&xFI$=W(wqUYs1i0o4!Qm)4R6D;&w^(Y>m<$7U0rT#0(rU&UzG48^ye%$c
z80(s6@Buh!pMlJ@`AS#NvhCse{#4$BmJGnCdm%lx+iv8}A4)r`xsCh^Y}D3(jT-LN
zs*?CY7j_l~FO3m6%=O+6_~nxTzkJGpAdHWF$}kYvfWb>{S`!`Wn9pHANDVu_X014Y
z<B||Zq>k0O{5%S<x^NNJ09Q~#8<Bdu-ld*i8db-7aLpc3tWuV}{fl2-Qax7^F&YN=
z<^6yxGTci7fCw{V-}MN9>=v-go?fEQ07Ye&I!Xy_)o^EMI)M*$z37fs;Z0j2Aa4!#
zmIN4|b&f7#&-j?Z2Ye8#N?{Ge33&B#%1-?R$2ID+cEkrZfpAI9T?Pj53AJY7OAL^q
z0OLc2f2uujLK}m63$WkDcff?8t5*4gMS$^bG9BQy0mfHUJXZ7D_zuQ}T(eu<mlN!O
z<g{=gISqFfR1%*Ix|X=!&1S6x$EhMDoB>s&_s+Zu&zn=a2fyg7_(0wu3lJBIIuI9<
zjD(`CZtdA$<XrRt8#C3~)F1e5fP1x2xQhk^hW3EmQvX0laMxuoFm&te9eC-FxE_EG
zg8N;F5&`JQ|FS<b<iq)gf-VTI2jGRN&!Q~Le}EXxe95v`L4<oo^=#Gmd;o^J?LX;h
zIKdy@PQbALY}P>c&9dA8L5$$sd-gE}q!EBR!M(b7^0;nb+1V@r#rk+W_^{ns$I4#C
zBcuRsiyGJlO|?i2`qfZL0(~POA~zN>Nm=&l0YVDESnpNkfM-x}53CI@MPu{Ra38in
zXstm9pi#e-1VU@L9BVAaFQ8yn(3-)zIG_N?d&8XuDv37;ZDhba7Swm(;$QnBnc2^e
z`u^^pBBBGA@ZlWrkQMQfkVh~h*<GM~$c{}Oz#b%z7Raj`wATtRZ_O=Y7OHkOfbE(a
z!`1?@U4z$d{^;HKy1n7R_H3)s{Pn1h2)p^N)%&6yBHZoT&$S4WY<vjB21rr^c5QnN
z>)=jq)@Ja%J=&<wNO+<!_4e^5q8R9P%^bxHi5}J1;rp4Rfi($t9&*8Q&gGfb)`*tT
zM$b`>t`)!ZUu1bvDQzlSkv<i{IxQr{lN-o4pM;S&8$w_zj2nr1;S^V_@{K9uTwYKh
z(fKDLGfc~>-|~wz`)f#A$-eSHKQ^_vJpXD%2wL*7O5FX1VDC#`r2692^H#o55#Ia!
z4dz+?GDo+2tXwWLC7us@toyAY?{PfhE=@e+eSi)kF<|xj%NftXDEr`F?;+CuF@uYK
z#;?CWhku&&&^r=6So1Dz%W00Y25I)bdCA1R{s)!47W0U`2bw2G!m8CEkeq)lnIY-h
zq9*fnsA4XSwVfDQRFb8x)8`gGXI3%e66ZYS#`3Bew2qexE#Z~B?=wA{Y~)AaY-BBy
zU1w64Va%VM1m>>#Po*=qHxXW_5Mv01p^SDCM{mn8|1|RtErsneg-ue$l}~&b40Iv&
z6{+QDf0VL{l<|%;E;>#%R68<h{!ybkLsX~QS2C%UmFIluoM^em+;svdDdC0EV%2H<
zbw{KAoqXsI1?_KCGC4~pyWQz}NLY1~Etv{o<&ZgPoyKeRu>`38A8}zlvdQ&)$x{o}
zH0gYuZ`ly^4uhwC8Z`~+dhZ>zP?X%NOSCZ@_L^@#a>SmwNVr9^3t3AXWJ(z6OMTia
z>TYCy-6^l_2)BZnYWypgE8Lm&yYr8KB(0WAEvUT;JCf%ga-~vWxtgbC$+p}su$aqA
zkM0*Ic{I(*z?Fcpow6lS0%1`JpJx;$ZLdu(Wyy%9nSKzBRSC-Bbo>y_2<c&PY{`nG
z4Y){MW8X)tlVb9smAKVJjv$tcWIXa%uL0f7RMP5T@2EhY&bDaagc@4l#~KKv38L=a
z;W!;LJ+AD&EZ3&7P-(sK2)RyyYtiK@m?w3BCDQ{|4Hz{9=4fUOKP$BNJ2_g-7XKW%
zLeViZn0js#{<(Y<;!+S<Meua`^QI`1rj@lYR2_#UjoNNXieM3y*h?1{_(-PpJK-k7
zAsz7oNQo8iofL;OJw9<Y1;JFY$gWiswZPY9t`@;Y+ktX*_}0}G(kM(7ywvOgDcvgG
zufoQJQ>vHrha7xG)pkFA>Ll$;NG{5s+sYRIk=4Gu$`P5SqHHg)5|yDI;u2FNmoi$D
zv7BjQT<ZFHW_P(Ohfa3AHd#0`xMxtA<cu9+9n^$#<6A-ZL!bfY4gmy0_50G@YZYc}
zXJ>_Vusj{zYrz=+%ltslRV;r;zw+*ux}vWHV?2+Q(fz0<?TK@!-lwHZP9`@EGv&f-
zhE=#H#fQqdx(!xRV;xQ<20g4~X)abH>*`_EBzgDKu|&V*EKtnq7S5~AV{}iL<$-ka
zUF(ZiFN+k`^#muZvB&WXwzMCI<ZFlZ91A<oZpZC`(EOAi#7W^pdjC2-vDHsFXH2*B
z1;aGO9d)JE0tF=|iu<1w;)2413@>7sODZ@UZ+z5JQ{|eduob8i$*P@lV_eo0GRj04
zb~JPZGrGtKt6(asG#4}3msq#ZRr=?QQjHO`t@bwewA0)QguTNmidD>-*Fr=4B8eBe
z+qSLUS+jo3isf+|N8F#YP?=$NN$Bso<QpF{eKdWfVTR3IK*)kZp@^_cn%K1}agKUJ
z(0|7cUusO!xJ!|73yN|!M&X~>!5exrkr80;_zbboe2;9Rw4+sm<;0XF$J=FL1}3>z
zxjx+EISt+ORN<YTr*&10T<6C=MYq@VmU;0_!*K1Aj&gx(2w0{KLI{NE_j3I<&3`W%
z)=WS4NmDimlOO$(rz8atBK&Ykkd};p0$<aCJ<TQ!MO-1Pec}lX0oh3{Q-k$ws0I~^
zvCu|Ed;ZGka8|myZo4gkf?GgAYOphE^=ZS#fIV#ESW;_emQqkU58iG<YWX*-tP5kT
z-cKz<(mEV0meoR7eTX0Dty;XIzLl50$mN(dJ`ig2aQwN(hH1fCw)*Dwy;sM#4Cu%-
zS$r+eD7}U#!8sF%IP{6TPeL?LDqV%7Wl{?gP#5@HuBy>2A}=yYb?VQ-b{*w^`ZyTM
z$8>%jAqhQ77^GF}1;L$=6L98HNx**Nh<^7<HMnGF>ttr@1iR^GZw5Eov*C}i&})DV
zKi*aC$Ml(#GZNdph3KOE<f)k+gvFh09z*TO+P3ley^|9=?S3@|<^7p^jq_=_4MbgH
z;CxcnaZOih{1DRorcXrnEWBQ&(Z3iOPPQ2HGH_gO9M_{8qdET}XW)v(Q*IvghwDq#
z<5qoR&|c!By;H@|d}LNGdHsyyvG>ixx$Hti@s~Pcp_DoozWCDiD9ydGRII%)M+zIH
zPS=TSH_KqSCk3?*RTyTf+d{l!Q-#G<LC_(w=xCNpM{*JP99By@PWDX+6|G%`x(ozj
zB1EiG&TL)^Df<4)+b$-#d9UjfsjsrTnOd|CGt^Sg2J>IfPJjo>dm2iJUhKAxFJIqG
zlHP`2yzGTn`JJo<3RAG%#1{enGJ+|5-=&=z-9*EDXjfAVa*c>M5*A$7NUp2l;nNZP
zy}dACeb^sgVD=#gnhk?TgPZ;S9t))7;357!`oFDz!Nb8VarcL}0*61OK`vMT+!FDR
zfFCI_R^ZSEiC`6Q&yIbSqT`3EU^{q#Rls25eU&bbLshVtW3URCx3#b0BXp<=mc9j6
z0oQ-;tBA`Ss)DTs2djW96!ukyt{$p_tyci6fK#;gRoJu-Rlz1~fmOh<gMF1Y<3m-j
zkp!>`_?7mL3ZA<?R%QPM7d#aF?yx^p9gg!R0Uii`7Vi%fclvYY|9vup2ZNss`-7*P
rv4j8eoB->9Z>ak^t1b}8A-C7-YJ@<q>b;iZ)DWH{5QvfM-lzWnUn`)c

-- 
2.24.1