From dada4a1355f50466ee69e0f5a08a8c00683d58d3 Mon Sep 17 00:00:00 2001 From: mhjensen Date: Sat, 12 Oct 2019 19:19:24 +0200 Subject: [PATCH] Dim red update --- doc/pub/DimRed/html/._DimRed-bs000.html | 38 +++---- doc/pub/DimRed/html/._DimRed-bs001.html | 36 ++++--- doc/pub/DimRed/html/._DimRed-bs002.html | 41 ++++---- doc/pub/DimRed/html/._DimRed-bs003.html | 101 +++++++++++-------- doc/pub/DimRed/html/._DimRed-bs004.html | 76 +++++++------- doc/pub/DimRed/html/._DimRed-bs005.html | 72 +++++++------ doc/pub/DimRed/html/._DimRed-bs006.html | 65 ++++++++---- doc/pub/DimRed/html/._DimRed-bs007.html | 53 +++++----- doc/pub/DimRed/html/._DimRed-bs008.html | 59 +++++------ doc/pub/DimRed/html/._DimRed-bs009.html | 65 ++++++++---- doc/pub/DimRed/html/DimRed-bs.html | 38 +++---- doc/pub/DimRed/html/DimRed-reveal.html | 77 +++++++++++--- doc/pub/DimRed/html/DimRed-solarized.html | 93 +++++++++++++---- doc/pub/DimRed/html/DimRed.html | 93 +++++++++++++---- doc/pub/DimRed/ipynb/DimRed.ipynb | 82 +++++++++++++-- doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz | Bin 191 -> 191 bytes doc/pub/DimRed/pdf/DimRed-minted.pdf | Bin 199386 -> 200327 bytes 17 files changed, 640 insertions(+), 349 deletions(-) diff --git a/doc/pub/DimRed/html/._DimRed-bs000.html b/doc/pub/DimRed/html/._DimRed-bs000.html index bbc53cf32..57a5f55b6 100644 --- a/doc/pub/DimRed/html/._DimRed-bs000.html +++ b/doc/pub/DimRed/html/._DimRed-bs000.html @@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source None, '___sec0'), ('Preprocessing our data', 2, None, '___sec1'), - ('Principal Component Analysis', 2, None, '___sec2'), - ('PCA and scikit-learn', 2, None, '___sec3'), - ('More on the PCA', 2, None, '___sec4'), - ('Incremental PCA', 2, None, '___sec5'), - ('Randomized PCA', 2, None, '___sec6'), - ('Kernel PCA', 2, None, '___sec7'), - ('LLE', 2, None, '___sec8'), - ('Other techniques', 2, None, '___sec9')]} + ('Simple preprocessing examples', 2, None, '___sec2'), + ('Principal Component Analysis', 2, None, '___sec3'), + ('PCA and scikit-learn', 2, None, '___sec4'), + ('More on the PCA', 2, None, '___sec5'), + ('Incremental PCA', 2, None, '___sec6'), + ('Randomized PCA', 2, None, '___sec7'), + ('Kernel PCA', 2, None, '___sec8'), + ('LLE', 2, None, '___sec9'), + ('Other techniques', 2, None, '___sec10')]} end of tocinfo --> @@ -93,14 +94,15 @@ MathJax.Hub.Config({ @@ -135,7 +137,7 @@ MathJax.Hub.Config({
[2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University

-

Oct 10, 2019

+

Oct 12, 2019


@@ -159,7 +161,7 @@ MathJax.Hub.Config({

  • 9
  • 10
  • ...
    • -
  • 11
    • +
  • 12
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs001.html b/doc/pub/DimRed/html/._DimRed-bs001.html index c7963e4f4..02f65d120 100644 --- a/doc/pub/DimRed/html/._DimRed-bs001.html +++ b/doc/pub/DimRed/html/._DimRed-bs001.html @@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source None, '___sec0'), ('Preprocessing our data', 2, None, '___sec1'), - ('Principal Component Analysis', 2, None, '___sec2'), - ('PCA and scikit-learn', 2, None, '___sec3'), - ('More on the PCA', 2, None, '___sec4'), - ('Incremental PCA', 2, None, '___sec5'), - ('Randomized PCA', 2, None, '___sec6'), - ('Kernel PCA', 2, None, '___sec7'), - ('LLE', 2, None, '___sec8'), - ('Other techniques', 2, None, '___sec9')]} + ('Simple preprocessing examples', 2, None, '___sec2'), + ('Principal Component Analysis', 2, None, '___sec3'), + ('PCA and scikit-learn', 2, None, '___sec4'), + ('More on the PCA', 2, None, '___sec5'), + ('Incremental PCA', 2, None, '___sec6'), + ('Randomized PCA', 2, None, '___sec7'), + ('Kernel PCA', 2, None, '___sec8'), + ('LLE', 2, None, '___sec9'), + ('Other techniques', 2, None, '___sec10')]} end of tocinfo --> @@ -93,14 +94,15 @@ MathJax.Hub.Config({ @@ -153,6 +155,8 @@ reduction techniques: the principal component analysis PCA, Kernel PCA, and Loca
  • 9
  • 10
  • 11
    • +
  • ...
    • +
  • 12
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs002.html b/doc/pub/DimRed/html/._DimRed-bs002.html index a720c0ad3..ca0c8c84d 100644 --- a/doc/pub/DimRed/html/._DimRed-bs002.html +++ b/doc/pub/DimRed/html/._DimRed-bs002.html @@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source None, '___sec0'), ('Preprocessing our data', 2, None, '___sec1'), - ('Principal Component Analysis', 2, None, '___sec2'), - ('PCA and scikit-learn', 2, None, '___sec3'), - ('More on the PCA', 2, None, '___sec4'), - ('Incremental PCA', 2, None, '___sec5'), - ('Randomized PCA', 2, None, '___sec6'), - ('Kernel PCA', 2, None, '___sec7'), - ('LLE', 2, None, '___sec8'), - ('Other techniques', 2, None, '___sec9')]} + ('Simple preprocessing examples', 2, None, '___sec2'), + ('Principal Component Analysis', 2, None, '___sec3'), + ('PCA and scikit-learn', 2, None, '___sec4'), + ('More on the PCA', 2, None, '___sec5'), + ('Incremental PCA', 2, None, '___sec6'), + ('Randomized PCA', 2, None, '___sec7'), + ('Kernel PCA', 2, None, '___sec8'), + ('LLE', 2, None, '___sec9'), + ('Other techniques', 2, None, '___sec10')]} end of tocinfo --> @@ -93,14 +94,15 @@ MathJax.Hub.Config({ @@ -123,15 +125,15 @@ MathJax.Hub.Config({

    Before we proceed however, we will discuss how to preprocess our -data. Till now and in connection with project 1 not met so many cases +data. Till now and in connection with our previous examples we have not met so many cases where we are too sensitive to the scaling of our data. Normally the data may need a rescaling and/or may be sensitive to extreme values. Scaling the data renders our inputs much more suitable for the -algorithms we want to emply. +algorithms we want to employ.

    Scikit-Learn has several functions which allow us to rescale the data, normally resulting in much better results in terms of various accuracy scores. The StandardScaler function in Scikit-Learn ensures that for each feature/predictor we study the mean value is zero and the variance is zero (every column in the design/feature matrix). -This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scale each column of the design matrix so that +This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scales each column of the design matrix by its Euclidean norm.

    @@ -154,6 +156,7 @@ This scaling has the drawback that it does not ensure that we have a particular

  • 9
  • 10
  • 11
    • +
  • 12
  • »
    • diff --git a/doc/pub/DimRed/html/._DimRed-bs003.html b/doc/pub/DimRed/html/._DimRed-bs003.html index c0ba3abc6..a64e6b72c 100644 --- a/doc/pub/DimRed/html/._DimRed-bs003.html +++ b/doc/pub/DimRed/html/._DimRed-bs003.html @@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source None, '___sec0'), ('Preprocessing our data', 2, None, '___sec1'), - ('Principal Component Analysis', 2, None, '___sec2'), - ('PCA and scikit-learn', 2, None, '___sec3'), - ('More on the PCA', 2, None, '___sec4'), - ('Incremental PCA', 2, None, '___sec5'), - ('Randomized PCA', 2, None, '___sec6'), - ('Kernel PCA', 2, None, '___sec7'), - ('LLE', 2, None, '___sec8'), - ('Other techniques', 2, None, '___sec9')]} + ('Simple preprocessing examples', 2, None, '___sec2'), + ('Principal Component Analysis', 2, None, '___sec3'), + ('PCA and scikit-learn', 2, None, '___sec4'), + ('More on the PCA', 2, None, '___sec5'), + ('Incremental PCA', 2, None, '___sec6'), + ('Randomized PCA', 2, None, '___sec7'), + ('Kernel PCA', 2, None, '___sec8'), + ('LLE', 2, None, '___sec9'), + ('Other techniques', 2, None, '___sec10')]} end of tocinfo --> @@ -93,14 +94,15 @@ MathJax.Hub.Config({ @@ -116,38 +118,54 @@ MathJax.Hub.Config({ -

    Principal Component Analysis

    -
    -
    -

    -Principal Component Analysis (PCA) is by far the most popular dimensionality reduction algorithm. -First it identifies the hyperplane that lies closest to the data, and then it projects the data onto it. +

    Simple preprocessing examples

    -The following Python code uses NumPy’s svd() function to obtain all the principal components of the -training set, then extracts the first two principal components +We show here how we can use a simple regression case (our nuclear binding energies discussed earlier). +Rescaling our data with different +

    -

    X_centered = X - X.mean(axis=0)
-U, s, V = np.linalg.svd(X_centered)
-c1 = V.T[:, 0]
-c2 = V.T[:, 1]
-
    -

    -PCA assumes that the dataset is centered around the origin. Scikit-Learn’s PCA classes take care of centering -the data for you. However, if you implement PCA yourself (as in the preceding example), or if you use other libraries, don’t -forget to center the data first. +

    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+from sklearn.svm import SVC
+cancer = load_breast_cancer()
 
-
    
-Once you have identified all the principal components, you can reduce the dimensionality of the dataset
-down to \( d \) dimensions by projecting it onto the hyperplane defined by the first \( d \) principal components.
-Selecting this hyperplane ensures that the projection will preserve as much variance as possible. 
-
    
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
 
-
-
    W2 = V.T[:, :2]
-X2D = X_centered.dot(W2)
+svm = SVC(C=100)
+svm.fit(X_train, y_train)
+print("Test set accuracy: {:.2f}".format(svm.score(X_test,y_test)))
+
+from sklearn.preprocessing import MinMaxScaler, StandardScaler
+
+scaler = MinMaxScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+print("Feature min values before scaling:\n {}".format(X_train.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train.max(axis=0)))
+
+print("Feature min values before scaling:\n {}".format(X_train_scaled.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train_scaled.max(axis=0)))
+
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
 
    
 
    
 
    
@@ -165,6 +183,7 @@ X2D = X_centered9
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/._DimRed-bs004.html b/doc/pub/DimRed/html/._DimRed-bs004.html
index e94811a79..9bd20ef9e 100644
--- a/doc/pub/DimRed/html/._DimRed-bs004.html
+++ b/doc/pub/DimRed/html/._DimRed-bs004.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -114,36 +116,41 @@ MathJax.Hub.Config({
 
     
     
     
     
 
 
-
+
 
-
    PCA and scikit-learn 
    
+
    Principal Component Analysis 
    
+
    
+
    
+
     
+Principal Component Analysis (PCA) is by far the most popular dimensionality reduction algorithm.
+First it identifies the hyperplane that lies closest to the data, and then it projects the data onto it.
 
 
    
-Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The
-following code applies PCA to reduce the dimensionality of the dataset down to two dimensions (note
-that it automatically takes care of centering the data):
+The following Python code uses NumPy’s svd() function to obtain all the principal components of the
+training set, then extracts the first two principal components
 
    
 
 
-
    from sklearn.decomposition import PCA
-pca = PCA(n_components = 2)
-X2D = pca.fit_transform(X)
+
    X_centered = X - X.mean(axis=0)
+U, s, V = np.linalg.svd(X_centered)
+c1 = V.T[:, 0]
+c2 = V.T[:, 1]
 
    
 
    
-After fitting the PCA transformer to the dataset, you can access the principal components using the
-components variable (note that it contains the PCs as horizontal vectors, so, for example, the first
-principal component is equal to 
+PCA assumes that the dataset is centered around the origin. Scikit-Learn’s PCA classes take care of centering
+the data for you. However, if you implement PCA yourself (as in the preceding example), or if you use other libraries, don’t
+forget to center the data first.
+
+
    
+Once you have identified all the principal components, you can reduce the dimensionality of the dataset
+down to \( d \) dimensions by projecting it onto the hyperplane defined by the first \( d \) principal components.
+Selecting this hyperplane ensures that the projection will preserve as much variance as possible. 
 
    
 
 
-
    pca.components_.T[:, 0]).
+
    W2 = V.T[:, :2]
+X2D = X_centered.dot(W2)
 
    
-
    
-Another very useful piece of information is the explained variance ratio of each principal component,
-available via the \( explained\_variance\_ratio \) variable. It indicates the proportion of the dataset’s
-variance that lies along the axis of each principal component. 
-More material to come here.
-
 
    
 
    
 
@@ -160,6 +167,7 @@ More material to come here.
   
  • 9
    • 
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/._DimRed-bs005.html b/doc/pub/DimRed/html/._DimRed-bs005.html
index 5d9967a80..fea5bd2b7 100644
--- a/doc/pub/DimRed/html/._DimRed-bs005.html
+++ b/doc/pub/DimRed/html/._DimRed-bs005.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -114,33 +116,36 @@ MathJax.Hub.Config({
 
     
     
     
     
 
 
-
+
 
-
    More on the PCA 
    
-Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to
-choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%).
-Unless, of course, you are reducing dimensionality for data visualization — in that case you will
-generally want to reduce the dimensionality down to 2 or 3.
-The following code computes PCA without reducing dimensionality, then computes the minimum number
-of dimensions required to preserve 95% of the training set’s variance:
+
    PCA and scikit-learn 
    
+
+
    
+Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The
+following code applies PCA to reduce the dimensionality of the dataset down to two dimensions (note
+that it automatically takes care of centering the data):
 
    
 
 
-
    pca = PCA()
-pca.fit(X)
-cumsum = np.cumsum(pca.explained_variance_ratio_)
-d = np.argmax(cumsum >= 0.95) + 1
+
    from sklearn.decomposition import PCA
+pca = PCA(n_components = 2)
+X2D = pca.fit_transform(X)
 
    
 
    
-You could then set \( n\_components=d \) and run PCA again. However, there is a much better option: instead
-of specifying the number of principal components you want to preserve, you can set \( n\_components \) to be
-a float between 0.0 and 1.0, indicating the ratio of variance you wish to preserve:
+After fitting the PCA transformer to the dataset, you can access the principal components using the
+components variable (note that it contains the PCs as horizontal vectors, so, for example, the first
+principal component is equal to 
 
    
 
 
-
    pca = PCA(n_components=0.95)
-X_reduced = pca.fit_transform(X)
+
    pca.components_.T[:, 0]).
 
    
+
    
+Another very useful piece of information is the explained variance ratio of each principal component,
+available via the \( explained\_variance\_ratio \) variable. It indicates the proportion of the dataset’s
+variance that lies along the axis of each principal component. 
+More material to come here.
+
 
    
 
    
 
@@ -157,6 +162,7 @@ X_reduced = pca
   
  • 9
    • 
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/._DimRed-bs006.html b/doc/pub/DimRed/html/._DimRed-bs006.html
index 8376feceb..93c298869 100644
--- a/doc/pub/DimRed/html/._DimRed-bs006.html
+++ b/doc/pub/DimRed/html/._DimRed-bs006.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -116,13 +118,31 @@ MathJax.Hub.Config({
 
 
 
-
    Incremental PCA 
    
-One problem with the preceding implementation of PCA is that it requires the whole training set to fit in
-memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have
-been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch
-at a time. This is useful for large training sets, and also to apply PCA online (i.e., on the fly, as new
-instances arrive).
+
    More on the PCA 
    
+Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to
+choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%).
+Unless, of course, you are reducing dimensionality for data visualization — in that case you will
+generally want to reduce the dimensionality down to 2 or 3.
+The following code computes PCA without reducing dimensionality, then computes the minimum number
+of dimensions required to preserve 95% of the training set’s variance:
+
    
 
+
+
    pca = PCA()
+pca.fit(X)
+cumsum = np.cumsum(pca.explained_variance_ratio_)
+d = np.argmax(cumsum >= 0.95) + 1
+
    
+
    
+You could then set \( n\_components=d \) and run PCA again. However, there is a much better option: instead
+of specifying the number of principal components you want to preserve, you can set \( n\_components \) to be
+a float between 0.0 and 1.0, indicating the ratio of variance you wish to preserve:
+
    
+
+
+
    pca = PCA(n_components=0.95)
+X_reduced = pca.fit_transform(X)
+
    
 
    
 
    
 
@@ -139,6 +159,7 @@ instances arrive).
   
  • 9
    • 
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/._DimRed-bs007.html b/doc/pub/DimRed/html/._DimRed-bs007.html
index a064b2321..3d7461c0c 100644
--- a/doc/pub/DimRed/html/._DimRed-bs007.html
+++ b/doc/pub/DimRed/html/._DimRed-bs007.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -116,18 +118,12 @@ MathJax.Hub.Config({
 
 
 
-
    Randomized PCA 
    
-
-
    
-Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic
-algorithm that quickly finds an approximation of the first d principal components. Its computational
-complexity is \( O(m \times d^2)+O(d^3) \), instead of \( O(m \times n^2) + O(n^3) \), so it is dramatically faster than the
-previous algorithms when \( d \) is much smaller than \( n \).
-
-
    
-
    
-
    
-
+
    Incremental PCA 
    
+One problem with the preceding implementation of PCA is that it requires the whole training set to fit in
+memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have
+been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch
+at a time. This is useful for large training sets, and also to apply PCA online (i.e., on the fly, as new
+instances arrive).
 
 
    
 
    
@@ -145,6 +141,7 @@ previous algorithms when \( d \) is much smaller than \( n \).
   
  • 9
    • 
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/._DimRed-bs008.html b/doc/pub/DimRed/html/._DimRed-bs008.html
index cb888f35f..c26ea557e 100644
--- a/doc/pub/DimRed/html/._DimRed-bs008.html
+++ b/doc/pub/DimRed/html/._DimRed-bs008.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -116,28 +118,14 @@ MathJax.Hub.Config({
 
 
 
-
    Kernel PCA 
    
-
    
-
    
-
     
+
    Randomized PCA 
    
 
 
    
-The kernel trick is a mathematical technique that implicitly maps instances into a
-very high-dimensional space (called the feature space), enabling nonlinear classification and regression
-with Support Vector Machines. Recall that a linear decision boundary in the high-dimensional feature
-space corresponds to a complex nonlinear decision boundary in the original space.
-It turns out that the same trick can be applied to PCA, making it possible to perform complex nonlinear
-projections for dimensionality reduction. This is called Kernel PCA (kPCA). It is often good at
-preserving clusters of instances after projection, or sometimes even unrolling datasets that lie close to a
-twisted manifold.
-For example, the following code uses Scikit-Learn’s KernelPCA class to perform kPCA with an
-
    
+Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic
+algorithm that quickly finds an approximation of the first d principal components. Its computational
+complexity is \( O(m \times d^2)+O(d^3) \), instead of \( O(m \times n^2) + O(n^3) \), so it is dramatically faster than the
+previous algorithms when \( d \) is much smaller than \( n \).
 
-
-
    from sklearn.decomposition import KernelPCA
-rbf_pca = KernelPCA(n_components = 2, kernel="rbf", gamma=0.04)
-X_reduced = rbf_pca.fit_transform(X)
-
    
 
    
 
    
 
    
@@ -159,6 +147,7 @@ X_reduced = rbf_pca9
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/._DimRed-bs009.html b/doc/pub/DimRed/html/._DimRed-bs009.html
index e95aeda9c..f6dd448e3 100644
--- a/doc/pub/DimRed/html/._DimRed-bs009.html
+++ b/doc/pub/DimRed/html/._DimRed-bs009.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -116,14 +118,32 @@ MathJax.Hub.Config({
 
 
 
-
    LLE 
    
+
    Kernel PCA 
    
+
    
+
    
+
     
 
 
    
-Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction
-(NLDR) technique. It is a Manifold Learning technique that does not rely on projections like the previous
-algorithms. In a nutshell, LLE works by first measuring how each training instance linearly relates to its
-closest neighbors (c.n.), and then looking for a low-dimensional representation of the training set where
-these local relationships are best preserved (more details shortly).
+The kernel trick is a mathematical technique that implicitly maps instances into a
+very high-dimensional space (called the feature space), enabling nonlinear classification and regression
+with Support Vector Machines. Recall that a linear decision boundary in the high-dimensional feature
+space corresponds to a complex nonlinear decision boundary in the original space.
+It turns out that the same trick can be applied to PCA, making it possible to perform complex nonlinear
+projections for dimensionality reduction. This is called Kernel PCA (kPCA). It is often good at
+preserving clusters of instances after projection, or sometimes even unrolling datasets that lie close to a
+twisted manifold.
+For example, the following code uses Scikit-Learn’s KernelPCA class to perform kPCA with an
+
    
+
+
+
    from sklearn.decomposition import KernelPCA
+rbf_pca = KernelPCA(n_components = 2, kernel="rbf", gamma=0.04)
+X_reduced = rbf_pca.fit_transform(X)
+
    
+
    
+
    
+
    
+
 
 
    
 
    
@@ -141,6 +161,7 @@ these local relationships are best preserved (more details shortly).
   
  • 9
    • 
   
  • 10
    • 
   
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/DimRed-bs.html b/doc/pub/DimRed/html/DimRed-bs.html
index bbc53cf32..57a5f55b6 100644
--- a/doc/pub/DimRed/html/DimRed-bs.html
+++ b/doc/pub/DimRed/html/DimRed-bs.html
@@ -46,14 +46,15 @@ Automatically generated HTML file from DocOnce source
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -93,14 +94,15 @@ MathJax.Hub.Config({
         
       
@@ -135,7 +137,7 @@ MathJax.Hub.Config({
 
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
    
 
    
 
    
-
    Oct 10, 2019
     
+
    Oct 12, 2019
     
 
    
 
    
 
@@ -159,7 +161,7 @@ MathJax.Hub.Config({
   
  • 9
    • 
   
  • 10
    • 
   
  • ...
    • 
-  
  • 11
    • 
+  
  • 12
    • 
   
  • »
    • 
 
 
diff --git a/doc/pub/DimRed/html/DimRed-reveal.html b/doc/pub/DimRed/html/DimRed-reveal.html
index fadbd8544..1cb7cc4d2 100644
--- a/doc/pub/DimRed/html/DimRed-reveal.html
+++ b/doc/pub/DimRed/html/DimRed-reveal.html
@@ -148,7 +148,7 @@ MathJax.Hub.Config({
 
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
    
 
    
 
     
    
-
    Oct 10, 2019
     
+
    Oct 12, 2019
     
 
    
 
    
 
@@ -184,15 +184,15 @@ reduction techniques: the principal component analysis PCA, Kernel PCA, and Loca
 
 
    
 Before we proceed however, we will discuss how to preprocess our
-data. Till now and in connection with project 1 not met so many cases
+data. Till now and in connection with our previous examples we have not met so many cases
 where we are too sensitive to the scaling of our data. Normally the
 data may need a rescaling and/or may be sensitive to extreme
 values. Scaling the data renders our inputs much more suitable for the
-algorithms we want to emply.
+algorithms we want to employ.
 
 
    
 Scikit-Learn has several functions which allow us to rescale the data, normally resulting in much better results in terms of various accuracy scores.  The StandardScaler function in Scikit-Learn ensures that for each feature/predictor we study the mean value is zero and the variance is zero (every column in the design/feature matrix).
-This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scale each column of the design matrix so that
+This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scales each column of the design matrix by its Euclidean norm.
 
 
 
    
@@ -200,7 +200,60 @@ This scaling has the drawback that it does not ensure that we have a particular
 
 
 
    
-
    Principal Component Analysis 
    
+
    Simple preprocessing examples 
    
+
+
    
+We show here how we can use a simple regression case (our nuclear binding energies discussed earlier).
+Rescaling our data with different
+
+
    
+
+
+
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+from sklearn.svm import SVC
+cancer = load_breast_cancer()
+
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+
+svm = SVC(C=100)
+svm.fit(X_train, y_train)
+print("Test set accuracy: {:.2f}".format(svm.score(X_test,y_test)))
+
+from sklearn.preprocessing import MinMaxScaler, StandardScaler
+
+scaler = MinMaxScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+print("Feature min values before scaling:\n {}".format(X_train.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train.max(axis=0)))
+
+print("Feature min values before scaling:\n {}".format(X_train_scaled.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train_scaled.max(axis=0)))
+
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
    
+
    
+
+
+
    
+
    Principal Component Analysis 
    
 
    
 
 
    
@@ -237,7 +290,7 @@ X2D = X_centered.dot(W2)
 
 
 
    
-
    PCA and scikit-learn 
    
+
    PCA and scikit-learn 
    
 
 
    
 Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The
@@ -268,7 +321,7 @@ More material to come here.
 
 
 
    
-
    More on the PCA 
    
+
    More on the PCA 
    
 Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to
 choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%).
 Unless, of course, you are reducing dimensionality for data visualization — in that case you will
@@ -297,7 +350,7 @@ X_reduced = pca.fit_transform(X)
 
 
 
    
-
    Incremental PCA 
    
+
    Incremental PCA 
    
 One problem with the preceding implementation of PCA is that it requires the whole training set to fit in
 memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have
 been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch
@@ -307,7 +360,7 @@ instances arrive).
 
 
 
    
-
    Randomized PCA 
    
+
    Randomized PCA 
    
 
 
    
 Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic
@@ -321,7 +374,7 @@ previous algorithms when \( d \) is much smaller than \( n \).
 
 
 
    
-
    Kernel PCA 
    
+
    Kernel PCA 
    
 
    
 
 
    
@@ -347,7 +400,7 @@ X_reduced = rbf_pca.fit_transform(X)
 
 
 
    
-
    LLE 
    
+
    LLE 
    
 
 
    
 Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction
@@ -359,7 +412,7 @@ these local relationships are best preserved (more details shortly).
 
 
 
    
-
    Other techniques 
    
+
    Other techniques 
    
 
 
    
 There are many other dimensionality reduction techniques, several of which are available in Scikit-Learn.
diff --git a/doc/pub/DimRed/html/DimRed-solarized.html b/doc/pub/DimRed/html/DimRed-solarized.html
index 79620e5a8..599608524 100644
--- a/doc/pub/DimRed/html/DimRed-solarized.html
+++ b/doc/pub/DimRed/html/DimRed-solarized.html
@@ -66,14 +66,15 @@ div { text-align: justify; text-justify: inter-word; }
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -115,7 +116,7 @@ MathJax.Hub.Config({
 
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
    
 
    
 
    
-
    Oct 10, 2019
     
+
    Oct 12, 2019
     
 
    
 
    
 









    
@@ -150,15 +151,15 @@ reduction techniques: the principal component analysis PCA, Kernel PCA, and Loca
 
 
    
 Before we proceed however, we will discuss how to preprocess our
-data. Till now and in connection with project 1 not met so many cases
+data. Till now and in connection with our previous examples we have not met so many cases
 where we are too sensitive to the scaling of our data. Normally the
 data may need a rescaling and/or may be sensitive to extreme
 values. Scaling the data renders our inputs much more suitable for the
-algorithms we want to emply.
+algorithms we want to employ.
 
 
    
 Scikit-Learn has several functions which allow us to rescale the data, normally resulting in much better results in terms of various accuracy scores.  The StandardScaler function in Scikit-Learn ensures that for each feature/predictor we study the mean value is zero and the variance is zero (every column in the design/feature matrix).
-This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scale each column of the design matrix so that
+This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scales each column of the design matrix by its Euclidean norm.
 
 
 
    
@@ -167,7 +168,59 @@ This scaling has the drawback that it does not ensure that we have a particular
 
    
 









    
 
-
    Principal Component Analysis 
    
+
    Simple preprocessing examples 
    
+
+
    
+We show here how we can use a simple regression case (our nuclear binding energies discussed earlier).
+Rescaling our data with different
+
+
    
+
+
+
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+from sklearn.svm import SVC
+cancer = load_breast_cancer()
+
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+
+svm = SVC(C=100)
+svm.fit(X_train, y_train)
+print("Test set accuracy: {:.2f}".format(svm.score(X_test,y_test)))
+
+from sklearn.preprocessing import MinMaxScaler, StandardScaler
+
+scaler = MinMaxScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+print("Feature min values before scaling:\n {}".format(X_train.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train.max(axis=0)))
+
+print("Feature min values before scaling:\n {}".format(X_train_scaled.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train_scaled.max(axis=0)))
+
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
    
+
    
+









    
+
+
    Principal Component Analysis 
    
 
    
 
 
    
@@ -203,7 +256,7 @@ X2D = X_centered.dot(W2)
 
    
 
 
-
    PCA and scikit-learn 
    
+
    PCA and scikit-learn 
    
 
 
    
 Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The
@@ -234,7 +287,7 @@ More material to come here.
 
    
 









    
 
-
    More on the PCA 
    
+
    More on the PCA 
    
 Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to
 choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%).
 Unless, of course, you are reducing dimensionality for data visualization — in that case you will
@@ -262,7 +315,7 @@ X_reduced = pca.fit_transform(X)
 
    
 









    
 
-
    Incremental PCA 
    
+
    Incremental PCA 
    
 One problem with the preceding implementation of PCA is that it requires the whole training set to fit in
 memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have
 been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch
@@ -272,7 +325,7 @@ instances arrive).
 
    
 









    
 
-
    Randomized PCA 
    
+
    Randomized PCA 
    
 
 
    
 Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic
@@ -287,7 +340,7 @@ previous algorithms when \( d \) is much smaller than \( n \).
 
    
 









    
 
-
    Kernel PCA 
    
+
    Kernel PCA 
    
 
    
 
 
    
@@ -316,7 +369,7 @@ X_reduced = rbf_pca.fit_transform(X)
 
    
 









    
 
-
    LLE 
    
+
    LLE 
    
 
 
    
 Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction
@@ -328,7 +381,7 @@ these local relationships are best preserved (more details shortly).
 
    
 









    
 
-
    Other techniques 
    
+
    Other techniques 
    
 
 
    
 There are many other dimensionality reduction techniques, several of which are available in Scikit-Learn.
diff --git a/doc/pub/DimRed/html/DimRed.html b/doc/pub/DimRed/html/DimRed.html
index 426a673ac..98ec982c9 100644
--- a/doc/pub/DimRed/html/DimRed.html
+++ b/doc/pub/DimRed/html/DimRed.html
@@ -71,14 +71,15 @@ div { text-align: justify; text-justify: inter-word; }
                None,
                '___sec0'),
               ('Preprocessing our data', 2, None, '___sec1'),
-              ('Principal Component Analysis', 2, None, '___sec2'),
-              ('PCA and scikit-learn', 2, None, '___sec3'),
-              ('More on the PCA', 2, None, '___sec4'),
-              ('Incremental PCA', 2, None, '___sec5'),
-              ('Randomized PCA', 2, None, '___sec6'),
-              ('Kernel PCA', 2, None, '___sec7'),
-              ('LLE', 2, None, '___sec8'),
-              ('Other techniques', 2, None, '___sec9')]}
+              ('Simple preprocessing examples', 2, None, '___sec2'),
+              ('Principal Component Analysis', 2, None, '___sec3'),
+              ('PCA and scikit-learn', 2, None, '___sec4'),
+              ('More on the PCA', 2, None, '___sec5'),
+              ('Incremental PCA', 2, None, '___sec6'),
+              ('Randomized PCA', 2, None, '___sec7'),
+              ('Kernel PCA', 2, None, '___sec8'),
+              ('LLE', 2, None, '___sec9'),
+              ('Other techniques', 2, None, '___sec10')]}
 end of tocinfo -->
 
 
@@ -120,7 +121,7 @@ MathJax.Hub.Config({
 
    [2] Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University
    
 
    
 
    
-
    Oct 10, 2019
     
+
    Oct 12, 2019
     
 
    
 
    
 









    
@@ -155,15 +156,15 @@ reduction techniques: the principal component analysis PCA, Kernel PCA, and Loca
 
 
    
 Before we proceed however, we will discuss how to preprocess our
-data. Till now and in connection with project 1 not met so many cases
+data. Till now and in connection with our previous examples we have not met so many cases
 where we are too sensitive to the scaling of our data. Normally the
 data may need a rescaling and/or may be sensitive to extreme
 values. Scaling the data renders our inputs much more suitable for the
-algorithms we want to emply.
+algorithms we want to employ.
 
 
    
 Scikit-Learn has several functions which allow us to rescale the data, normally resulting in much better results in terms of various accuracy scores.  The StandardScaler function in Scikit-Learn ensures that for each feature/predictor we study the mean value is zero and the variance is zero (every column in the design/feature matrix).
-This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scale each column of the design matrix so that
+This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in Scikit-Learn is the MinMaxScaler which ensures that all features are exactly between \( 0 \) and \( 1 \). The Normalizer function scales each column of the design matrix by its Euclidean norm.
 
 
 
    
@@ -172,7 +173,59 @@ This scaling has the drawback that it does not ensure that we have a particular
 
    
 









    
 
-
    Principal Component Analysis 
    
+
    Simple preprocessing examples 
    
+
+
    
+We show here how we can use a simple regression case (our nuclear binding energies discussed earlier).
+Rescaling our data with different
+
+
    
+
+
+
    import matplotlib.pyplot as plt
+import numpy as np
+from sklearn.model_selection import  train_test_split 
+from sklearn.datasets import load_breast_cancer
+from sklearn.svm import SVC
+cancer = load_breast_cancer()
+
+X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)
+print(X_train.shape)
+print(X_test.shape)
+
+svm = SVC(C=100)
+svm.fit(X_train, y_train)
+print("Test set accuracy: {:.2f}".format(svm.score(X_test,y_test)))
+
+from sklearn.preprocessing import MinMaxScaler, StandardScaler
+
+scaler = MinMaxScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+print("Feature min values before scaling:\n {}".format(X_train.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train.max(axis=0)))
+
+print("Feature min values before scaling:\n {}".format(X_train_scaled.min(axis=0)))
+print("Feature max values before scaling:\n {}".format(X_train_scaled.max(axis=0)))
+
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
+scaler = StandardScaler()
+scaler.fit(X_train)
+X_train_scaled = scaler.transform(X_train)
+X_test_scaled = scaler.transform(X_test)
+
+svm.fit(X_train_scaled, y_train)
+print("Test set accuracy scaled data: {:.2f}".format(svm.score(X_test_scaled,y_test)))
+
    
+
    
+









    
+
+
    Principal Component Analysis 
    
 
    
 
 
    
@@ -208,7 +261,7 @@ X2D = X_centered
 
 
-
    PCA and scikit-learn 
    
+
    PCA and scikit-learn 
    
 
 
    
 Scikit-Learn’s PCA class implements PCA using SVD decomposition just like we did before. The
@@ -239,7 +292,7 @@ More material to come here.
 
    
 









    
 
-
    More on the PCA 
    
+
    More on the PCA 
    
 Instead of arbitrarily choosing the number of dimensions to reduce down to, it is generally preferable to
 choose the number of dimensions that add up to a sufficiently large portion of the variance (e.g., 95%).
 Unless, of course, you are reducing dimensionality for data visualization — in that case you will
@@ -267,7 +320,7 @@ X_reduced = pca
 
    
 









    
 
-
    Incremental PCA 
    
+
    Incremental PCA 
    
 One problem with the preceding implementation of PCA is that it requires the whole training set to fit in
 memory in order for the SVD algorithm to run. Fortunately, Incremental PCA (IPCA) algorithms have
 been developed: you can split the training set into mini-batches and feed an IPCA algorithm one minibatch
@@ -277,7 +330,7 @@ instances arrive).
 
    
 









    
 
-
    Randomized PCA 
    
+
    Randomized PCA 
    
 
 
    
 Scikit-Learn offers yet another option to perform PCA, called Randomized PCA. This is a stochastic
@@ -292,7 +345,7 @@ previous algorithms when \( d \) is much smaller than \( n \).
 
    
 









    
 
-
    Kernel PCA 
    
+
    Kernel PCA 
    
 
    
 
 
    
@@ -321,7 +374,7 @@ X_reduced = rbf_pcaLLE 
+
    LLE 
    
 
 
    
 Locally Linear Embedding (LLE) is another very powerful nonlinear dimensionality reduction
@@ -333,7 +386,7 @@ these local relationships are best preserved (more details shortly).
 
    
 









    
 
-
    Other techniques 
    
+
    Other techniques 
    
 
 
    
 There are many other dimensionality reduction techniques, several of which are available in Scikit-Learn.
diff --git a/doc/pub/DimRed/ipynb/DimRed.ipynb b/doc/pub/DimRed/ipynb/DimRed.ipynb
index 2a1cdc58d..b9620502d 100644
--- a/doc/pub/DimRed/ipynb/DimRed.ipynb
+++ b/doc/pub/DimRed/ipynb/DimRed.ipynb
@@ -10,7 +10,7 @@
     "  \n",
     "**Morten Hjorth-Jensen**, Department of Physics, University of Oslo and Department of Physics and Astronomy and National Superconducting Cyclotron Laboratory, Michigan State University\n",
     "\n",
-    "Date: **Oct 10, 2019**\n",
+    "Date: **Oct 12, 2019**\n",
     "\n",
     "Copyright 1999-2019, Morten Hjorth-Jensen. Released under CC Attribution-NonCommercial 4.0 license\n",
     "\n",
@@ -34,19 +34,79 @@
     "## Preprocessing our data\n",
     "\n",
     "Before we proceed however, we will discuss how to preprocess our\n",
-    "data. Till now and in connection with project 1 not met so many cases\n",
+    "data. Till now and in connection with our previous examples we have not met so many cases\n",
     "where we are too sensitive to the scaling of our data. Normally the\n",
     "data may need a rescaling and/or may be sensitive to extreme\n",
     "values. Scaling the data renders our inputs much more suitable for the\n",
-    "algorithms we want to emply.\n",
+    "algorithms we want to employ.\n",
     "\n",
     "**Scikit-Learn** has several functions which allow us to rescale the data, normally resulting in much better results in terms of various accuracy scores.  The **StandardScaler** function in **Scikit-Learn** ensures that for each feature/predictor we study the mean value is zero and the variance is zero (every column in the design/feature matrix).\n",
-    "This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in **Scikit-Learn** is the **MinMaxScaler** which ensures that all features are exactly between $0$ and $1$. The **Normalizer** function scale each column of the design matrix so that\n",
+    "This scaling has the drawback that it does not ensure that we have a particular maximum or minumum in our data set. Another function included in **Scikit-Learn** is the **MinMaxScaler** which ensures that all features are exactly between $0$ and $1$. The **Normalizer** function scales each column of the design matrix by its Euclidean norm.\n",
     "\n",
     "\n",
     "\n",
     "\n",
+    "## Simple preprocessing examples\n",
     "\n",
+    "We show here how we can use a simple regression case (our nuclear binding energies discussed earlier).\n",
+    "Rescaling our data with different"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [],
+   "source": [
+    "%matplotlib inline\n",
+    "\n",
+    "import matplotlib.pyplot as plt\n",
+    "import numpy as np\n",
+    "from sklearn.model_selection import  train_test_split \n",
+    "from sklearn.datasets import load_breast_cancer\n",
+    "from sklearn.svm import SVC\n",
+    "cancer = load_breast_cancer()\n",
+    "\n",
+    "X_train, X_test, y_train, y_test = train_test_split(cancer.data,cancer.target,random_state=0)\n",
+    "print(X_train.shape)\n",
+    "print(X_test.shape)\n",
+    "\n",
+    "svm = SVC(C=100)\n",
+    "svm.fit(X_train, y_train)\n",
+    "print(\"Test set accuracy: {:.2f}\".format(svm.score(X_test,y_test)))\n",
+    "\n",
+    "from sklearn.preprocessing import MinMaxScaler, StandardScaler\n",
+    "\n",
+    "scaler = MinMaxScaler()\n",
+    "scaler.fit(X_train)\n",
+    "X_train_scaled = scaler.transform(X_train)\n",
+    "X_test_scaled = scaler.transform(X_test)\n",
+    "\n",
+    "print(\"Feature min values before scaling:\\n {}\".format(X_train.min(axis=0)))\n",
+    "print(\"Feature max values before scaling:\\n {}\".format(X_train.max(axis=0)))\n",
+    "\n",
+    "print(\"Feature min values before scaling:\\n {}\".format(X_train_scaled.min(axis=0)))\n",
+    "print(\"Feature max values before scaling:\\n {}\".format(X_train_scaled.max(axis=0)))\n",
+    "\n",
+    "\n",
+    "svm.fit(X_train_scaled, y_train)\n",
+    "print(\"Test set accuracy scaled data: {:.2f}\".format(svm.score(X_test_scaled,y_test)))\n",
+    "\n",
+    "scaler = StandardScaler()\n",
+    "scaler.fit(X_train)\n",
+    "X_train_scaled = scaler.transform(X_train)\n",
+    "X_test_scaled = scaler.transform(X_test)\n",
+    "\n",
+    "svm.fit(X_train_scaled, y_train)\n",
+    "print(\"Test set accuracy scaled data: {:.2f}\".format(svm.score(X_test_scaled,y_test)))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
     "## Principal Component Analysis\n",
     "Principal Component Analysis (PCA) is by far the most popular dimensionality reduction algorithm.\n",
     "First it identifies the hyperplane that lies closest to the data, and then it projects the data onto it.\n",
@@ -57,7 +117,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 2,
    "metadata": {
     "collapsed": false
    },
@@ -84,7 +144,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 3,
    "metadata": {
     "collapsed": false
    },
@@ -108,7 +168,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 4,
    "metadata": {
     "collapsed": false
    },
@@ -130,7 +190,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 5,
    "metadata": {
     "collapsed": false
    },
@@ -159,7 +219,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 6,
    "metadata": {
     "collapsed": false
    },
@@ -182,7 +242,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 7,
    "metadata": {
     "collapsed": false
    },
@@ -228,7 +288,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 8,
    "metadata": {
     "collapsed": false
    },
diff --git a/doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz b/doc/pub/DimRed/ipynb/ipynb-DimRed-src.tar.gz
index 0fe79a5ff926ca1e164bcad734b5bcc4f1723935..c7ace3921ecea930e40f0c7ef25dd4c6a3633686 100644
GIT binary patch
literal 191
zcmV;w06_mAiwFP=3!+^B1MSbv3c@f92k@Qu6nTP?x^_DY+rfh%!x!jS=c=xqZ3ph{
z-3RDN@iIi{@A4-kgrvx~**=fl-36;b#Dr3qGELZ=xLN8MAq+^&7*b@EP(TvLGz|dP
zJL#pjj+^n6<}8#I>h0V(R@Wc)EU&;b|HQG87IufhHClnvPLx_-Ad_t&sz`Q&DxuQC
tPAxFv)=MLRs{wdbNhh`9*RV7CWO!?%@Ym0Hp67XAdjKoz&G7&T005T^U0eVF

literal 191
zcmV;w06_mAiwFQU5}#cF1MSaE3c@fD1>mlGia9|^Op~sKcHu&h;ssKY+Ne!xl7hXx
z{Q+GmZi)!`HjglmFjE%Wd|xE~?xWQpgjkxwlo_LQqRsN2VN3xNnq~}RG!_shgt7?G
zdMmy3&T~7i(wwDsQorquwYvWB&vFVp^G_TbkD6fp>*SNFhWO?h7_}9k-K@bFAdjJ#3nEwC>003QySSuxDbvM*Nm6n{#1}9qAyXHiKT2Ves!1TcZ^;2I3dSgV#J
zG5Ey96vJ*qkKly~PxUU_GFe;p0I0V|H&6_QOw{+$=vw@ShcV_4~j7QLzGcyz6n30qB_#g+fu4)NpPI
z-p1)L@AFG89L5X`uZjvMrCbtPg}fA)9<<0B(dPpamg+yy1QG1(1r(QMNbpxo(^*?gIKv%o))GHo)oB==k~
zCRwI=m2hl?#Mvq0nFPE=7fsyi5o;InX;Puj>3-Ck5a{-9sPfA6oV|`z+e$va`r;Nn
zNwa{*NwV8Yd}G@T{;k{L6(kF$v~uOc8>n=4fu)0QXCY%iT6uuilTXq9B;3BjrM1A|
zNdBfSi(4n;)60M5O6#xn2@fxKAG6D&-aqh&1iR%!2CSproiQuN>2-csM$_H!;sTea
zNxxN}d^Mic2CQ)kejv>`^;*LnB?>P;*^+yuhdmv2{zD}}AR0EgI5UlVCV|!$gj82U
zi#);rd_Wwi1v0|UsPVNnxu=7y93n=7a}q7Dnk1oTaK5XOFiqWcl-GSMELR!pavve15yQu>1owDkUqEy
z)wTy)?AC8iwC`hK%&r8`#Nje8V%vi78ps6c2WVSe4*ZKLhMndZlm_(d?bu^gAcGQG>h*-Hgf|m-PHM^c{`;(G7YR0>^woXZ
zq6(TWOK!ya8#&WovOlqC)t(FTKWaj`kS^uZqtDOi;m1_m`5C&%@dVMV8*PPFUc#!x
zz^F+a!3PHR82rkq^9;>&F!ak;9jP!u3L5~AC7yFS)XiKubf!Y-AXhq9j`mx#0g$ki
zN0H0E3vm_@Qx(vwjGcS`%>HTn44)m?jSy5GXshK)F6#A7^hW&q$%H+jD8mXJQ8yxg0QhE3#h+RNM
z0=6&+ibZ*=sq!t9+NnHHQS?QgJs^NW0OsDWlYi!CbOcb
zmyOVb-6#3)CRRO4RYMOEJ!JYC+V!W`$~d=JNq7eQ
z?Pok{1K!&azAz-3?xt$ZWgsH)BrUA-3N5l0W)82>_WG{VJHAI)R4ql
zc&kw!-~ROr^ngHM7EVa~-m0+P>^F!76U_MT4~b!Y-)<$*-=!QOo^15|GfPysaya`q
z^&O>9xP=)v5==b|r-O@PkX+<%DW0=XwTgZ?rwcVy`A08XG*~OW*k0?pu!2NF9P-mb9g61uvUd{$oVgAKd{Ug@={G-iHm
zWD#qOFA4okWAibAI%^W%_`fcc%mh4bE{^kmA9$SaTWpdc^>KsJB-t6TOSyW-G4^oD
zw#!Zn#Mg#>mwFM4s3?55cA}b$`bal*$3}D1kVU>A#VL8<9=trwZK=RU<3YwVR=p+@
z4SBGWMxB<{FQ`s-DCeDhP7$o;28xDSNtqj5Q#Z#S=i&q9^JNz5pdd`X+efy@1H=Mb
zQ}&?8;a+$0)g%oME;;*Z6A=3w1U>(*^M5&L20cFcJHm`ijeFmE(l?4HIOGgg{Z<}0
zCUKwIH|kC5ht7|jtsMd^^>B?L)C4srm(AClN?zey2e#1g{*eWVI48%i__!OsDcb4n
z1l(drx3?aD+}$g^9}VK%IG+1(;7-XMbB&~m7GQv%4l?d#!jr!DUK#|0rsUfmyrOv%
zG_&NiqwqE-0ZfiGhu3E`_J$Gr@ZapC{l=?2MA!Yc$J%#c%ou=G4<1#iNOZmEws@WX
zEg=ZRqOb{H*s(qFeiDYUZHJpf-YYH_BVQthRsxYj?|4y4A)g
z*`o20`38S#7*c^wh{9zB4w=4AR{`<3c|9Tc)*)RfhQLOP4Bt%V$dMmj`K3D!=e`rN
zFj93PH;a4k5Js_`GfE8FD;wpX;eO{|GbRzwTmx
zMZc>XdmJ8QILujGB$7YJ!DM*TfW?!YB2;6iLH?s_joktwH@g{+6HH#9=5@vu4I^bU
znPvQHwB@Mk7T9PKAaE`m@;ev)^`dxBV|()NB;Vz4Yh=jkDByVY7#0JCmtxGygj3D!GtG&i7!eM&3Iqto3Q*u8R*OxiY}2ZlM!mF
z58mD7-0)c($C;)!hx5Df+NRpdjXQ=CxuXY?ED>f$h%_Ub%1$Q8?mODO?u9%uLVA+G
zVX445IXV9Ct`noJ;Ih$z+Id}*CO4*@GC&~`MBK`g=_Y91=`<+oJW{NR5s^@P6px-r
zru%+=kGHuKPbT{$0UiL-Hv8H34jSqH+-~_?a{c4=v;Wd#yF2P$ANNoCn0xcPK7pi4
zgu^W3)}Rd9k#^m0R1<;uOgE@Ca76o2oOlyJ)@M_GXl=eD!9{Zw{HV3
z&z=_2!n{$vHNRUw&!4Z8b;i(m%BB12)opl)ZQe%}xrEZK`bY8y7Q?cQRTEu%EdHpOy^@t
zdgJR=04TGH2wjbJXA_6=Q~rEB1g?qH@|-oQs}224;jDD&lVds!s3A`mD)>R{401TU
z!d2NM%l*o63T45e;lspoo30;H!^WWP^w@)j34a@yyW{skN`lsx{a`Up8}%4IT8WV~
zY=Cvw15s*sewvpp?6VBCu@W`J_l>vzbbKFn`~^`&b4-
zcG+^m@|w?{xZJAN&~uOQ@^)a~J+TVlU#zl|r?}_KV!-ebuH&ZXh?hTJ+brd$hs*k#
zFI*pPB6sIZNC53Z;N7^N$GLCW?nA^Cj}N`L3u9(wPY}I6q6yDuq>Sx_Q^b?GYp59t
zz)35oL+p!9Te=F2q1AmizJawIeTYE*Tef6}g!X=?~ACV94
z#_t2kL`T;GeW@!ejtm!EKkapRl1nBB%!Q--kmk8O|CtIe19-?B_Fa_1PobIXXgRbAW)_)IDv52z^jh{m;
ziHUzZg>r&v7*I%xS`J!%Hx&*9B1Z@DU`c-UjCMTO3!4NekAhNoq8Y~kzENYLJcP)Z
zOld#g1&;oO^crKh&Osl8zRlkUSh`8vY=vlKq!=gv);_ZU<@aW-mxc&}a7-4f=5J$Kk}u>oY2Av1lK0%?vf7}h(#h3O
zFi2r1tCD4+QH=pfqFXu!ME)36&y{XVP?Aq&p6VLIUW>?Otx*+}QX#r{59}70J@^v0
zKBInVz+gnX8Tqwm)N;h}lY9gXH#x{#>%jXsSS~CZt#vGEaaOf^nK{FKnhb+>kvcv9
z{EpmOjB+2y*!!GYXBodpBAyVRbwE`28D(%$0p90AJPIoR#PKJc%)sP&=Y`NMZ07Z9&ZTvALR);P^Nx?_N0}O)3tu
zk!jAaK^Cd?K+m2DJz%Wq*QqyFden&g3Otk^2P^-J?3BzHzx)rDWywt!j)C2uQeE^C
zQ;GyG3%mTAqge*dfJ0r1&tHUt7uBV;fh;X=YUCHSOhP9v26yOUK`}
zFLij+(1gyl5kSw4RYJ!kLDER@PD0;Mh}0;g^9mf9C1yOrDdO#gr!FB5+2=Ui_K)1R6_%aC(J
z-{M+E3Lda&9rwOhpW+Yn^nT2#ub^Y5YBYVJQN+MNKz`qSFCo%?RxOUMeo&yoFVPEg
z-S>bo)kjB?DsS7Z6&S@Y7vG)RL#$KWjg<(!f4+!p_My=Um7|+y=lf+q`jhE5j62ze
zrz3>OkoT+p9AesKMmV-xOT4P6V%6y%e*@|wh}7^X^aJV$v({e2Bn8?NLC?}$ngMIp
zZ4qiSAkO8%2K&>a>b7Xj$1dG}%HvO)MPT0GzV-}!yee+FJny`+N5yZR-koRAkXx>9
z{c#h$t@ASC780`XhdIaR|5k3*9Ej0WmwZhZKgaf}a{mrvPNL?5_}A@G^xNZqgqCng
zq945!OxpaZXYBDI@Sgm)AT7&Z3wINLIT*EO0aGwx)3_Ummv<>-JbbU%c<=(BSo^+>
z*lli>J+`S899#m2W`XD0Pq{1UB$P+lv58j|ISL!OqSQaMuXADtkxRnhjka{9y&qD#
zw}Qtds|rJ4yy<5bq6th=<+Un37QP0Q%si27wman$0N2xWKf9)X9~pao;?aWzoxnC5
zK+^~2_NDuLFqsw8eRm|S^cx%6)Ws=2o5s=oGz+3>6w(k#57~7`Ox!=X-j9Zpvhbmq
z?td-5a4d?vrK01F9V+rEo+pCP>+fJ!JB@-u2&r{9Ms*jwhcSsuQUtzp=(=<;
z%yOJAbHfeB44uLvwT7a;3o)D=l^hE6e9PdkTBLIO+u|*lK3I(}LQ&r0uJ{K5xnr7x
zVzm*k-<6{YaEA%uG8*%a`>uNMRY!9#6I52Xd$de(3RHz3l}$mE*C9=PW+|CLT*Nf+
zV$+PRFfk$9Sqh%S1}@DD-I=*zNjt>)QF#8HTb1t3-njgRWQ3weaM}ividlK$SW#R`
zy;1&wV^-gZ`N!}7B;N>?9+(nT4pMlqUk*yb3izF`U
zU-+&VNa%?ihY*N~fWPHyDS;odJwr{BT$hsjOHq2PUm)zXOo+cZfO<+Gu2U}9rSEJd
zuj_o8Z1Wv~Mx)67UEw-qBK>F_NpAq6mD69bxIpaC{VrsN8=eklW^YRA;_7T>Z1+Fw
z@XH#Gjhlp-xB%wYX-9)-to8I_#ffa6fNDFJc_B?(&W1>ebW!&RfJ3aeH~0yG=Iaek^Q^tI^-^
zM~08KSm?Fk$L_1|dR5~>kx0v;SiN=bh1D~EXnA-syI|;3oRv?v$^eTCiryLL_G^OL
zjrJ?Doc*-4;qF~z?uaGUq?Ss$00kfsccc@)>(XmypEowRpdzx^b-Tu)Dk2!DWK$j*
zQ5$!!5B=C%2=Q{QZ`p-}PZ~1L>4mF5=$^Js3*%h$
z2g56|(_?#Jou7yktgmLM%~?A5+0Hx
ze87r4!Ola@*Tv6wrWx=aLMz`;1xDR=A~=7TxW{V#?4&;k^R=UQdU`6TYJ^i5tNg8t
z1ny;?2zlub5)S@8Op91vOsp3;K}q9-HgWRYQCvxcd?6)l`W&Q<$l`c**Kqh=Z9Qgf
zxF9cS)637@0?Opx=zHRI1`6=)_)|+^9C@*#50q=59nSn#G*mtaVkFSFZzbLta1?rR
z4ghoTJP9DYv%z7C3M*UA)}W%C5KfI!`@WuPTd`tf6OU0YdcE^x+5@+msl$Ps7Zj(o
zho23V$5FPQOWs$;f>yy9Wp$R#3mD$0J#
z2lXVr^z@OjeVT}0HP=VbZ+M)*4gW3
zB6*fIY17hVIw9Aq6iu=Eb5#O`^N$xyaV>o9vL#pv8c1m5zmh$05Q`L+{kOh60Qtsp
zR$$Ca3$ELC%3k;ElMf{P{!tb3y0yJcEp)hekiT?_VXI=(v(;2}sy1oDkC10L4}Gqg
zZ=*%3ftciOzw92)gQ#`I`C(ADQ+H_!53Ic#`tRfy_*?VMOxSxtg*4r$f<*S(@?z?D
z>s#Q(;tdt!==t>!ZOzdO;w7ugKyjea?Cy(UNL}YaSN$3rkO8JY?E>ju_t+$=L0<)3
zxzK)|{}f$=+RvH74m>9KrbUWL%nwjPiH3}Vg(wJS2uNVLI?{Z(XFBcdIK
zkm1l@P)!>MAfmvINc~U9XI*c*fRpt)k2o?N%KAMr;mWA1(b2uXAI#tf*RxHsQSzR$
zo6Y>PXToFthQu#w_SY5(BS*B22{?InXXObe!HO!Oj2Ht;oCM0nCm|xJdJ~?~nlK?Y
z8}PtRRJfDuiNvyo8j7eAd<|qlH2CZ^iC%lf*Z5EyMldj?H3asg#%ocxmAdG{v9Z-!
zAV$Lu!m_i~61$UegcE6txWSZxWen=klW!algzSz>dlq1*vm99F-a&Qa-fv>2vW&p6
zs0xM?tZl)>#^3j6i=v-DDmb9aSJ!c#40@*C1p8Atg;mf+w*w6W;TB$$M8TiZFaQ}r
zXL874T^eW)hl!aN)M)P;r_%~Dn~370af7P5x(EK||shm-QT#(5Y(Lk8M?5
zfF*I1=vmQQY_F1~c-V{!wuBkJ{fBZ7>k6@l^BUzmsBlj0I}NynI>bquaKj(3bX^gz
zwi!hhqHPyQddbAuH@rks(
zJY^v_*n01bAGDQf@6Ga$KT6zQ!@M6R?-4Bi`Zz{3XCMgVNH;sy0R;W^~}=h^b?>p+l;h6FAKQp%z-@dxEQU;yQm
z5^jkNQXTj}Eo(x&UP>;k4XY_*8^R-kq)O1zT$R(*tYsrs=Kby6FWk+03sh@mf3fK;CGKGp#l;7|62r_(vQv!EM8ef*pou>OACggIucI*AYaB&O!j0!G!#R#r
zh&N~=G#%~o4w;GPwDD@(sV1qFyUqh8D5Svt^_~}?4$UL$NMDx+k(TcSlxInoUznhv
zkW}((TnGzOnyWNNfTdBJmDT>jaSRpQuVQaugC(6qEY$LJQk^wVKhSqleU7&MC>H#^<~
z*mi3n>k`rXk8vMYvsmy7y%YF`RlBf$?Jj8zJu<13WtgM3qf{is>Jk!@Skp8&{B
z1s0672A63*D~4sH*I3fOevSyeS;!OgI(KBJ{+P5H
zLP$OyG?eLZ^q#B%QXsFcw)CjMvzDGQj4(0fy!K-U@0ek-?(N}Jt!~1;91<(KR*eWF
z9upFE`_om4Ga1EsD=^Z;$4mrsWS9m1MDQ_Q_ssm@ovJ>#EII~B%u$vaBdgvYjCr`&
zQ^nT)T;D_%g0D$1jt4;&4TcdJQ^F82j*e2wfLDjwVtJwmC=(-!_6of+<(p8^+F9r~
zYR#FzVLt%b_ed^kL^(~|8^thXp43!zcSdi}f7LmDrz)>P(&|&xnBOJ@K9mH#)T6My
zQcVn=9U>CZV%=_T3jgrLs>Rhtr0t&Rr37munIe*UDiY}8R$G67U9;*bYCf1q=>*3U
z*y{4%h4NJcH-jPAC0d<{hX$FlW#{d1XOg6h$cv=oR_L$Z$jxpfIP(R7&fA
zxtcdkQ~9&WXXDrK14`pnGY-;VQ;g9STgElzv4Pbk+kz*rNbI{d93*Xaa1*S3Wz-gqxPDBU>xi(QwQw6C(Bu-ez?uEWu+jd7Fq2oZ}Yk8m0r`f2J
zpH&SV{iBri&YXhF%>vF8~gwB`>iRLw9bjpb*y{2@Ii?b_8}p}p<{99P$)NEJfO-JM;>
zfLCGq)EojPxx5Y1TGDSHvqY(MYAsR`ZwP%|Wiye~5fe7#*kbJ>iI0$mANIr6gC+p!
z#M6~!CMcHPdfHkHHoi}L_+uG?h&Y|4e!0cDy`@ua?kx)SL0wrtVVvuI>R1YOUr#DAWC8=T53qz`IcxddIa%-~Hed
zd*@0l87Nryt8uxMRFBWXu60o_OHVh?%z;TkV)#0u@5{Z6G6W2ya4M!mVdF0uKo|v#
zDl7Zz)eE0ZH_Zs!uD+#S7Yx{+S|
zyqUwK^p~*e-0@6ib;0h{04`=KU^lM?f+f8#1}nFeMU$u4H5a`TvdCiTKvKm(L8iEo
zucM}{mB_eXz`gJFdp!28>A8@ezv;JgjsZEie3zNe6$3P?@*)*Df*9*eOOZqP0}{*h
zzLyCm6!f|`_>((GpHQZRC>u-+*OL#M_z!Ex&u7ZfI7b(~?I5?#bp^6yK&VG`2tQ5)
zPs~hle++@FjZ!&#V?alzJ=%PL8z<
zw=pWta^|x~7Xk>%$Bzu=)%MgRve9&uj0Zi;02D_rq?R_A76eeuTPk9UiNjbdD31_t
zf-M@?4jw7gpJQ!=^kvMGK+Y1$%sovbx>r-Jn&>K6&xoQ&K5F_>y1!|~zhqv2eCpjG
z1UmOip47qlfS*@ZcPC;Xe1h%9U~%1ezFqdx(hk2PRqE)d?vBL!5vF&dEY7f1e{(i#Q0MxGR-;l*k&x22)?v`I7@hf{h2#-Pxv0Ktgy7w-^$O@Gm7u+YPB+s
z$Pxc4r&@uLXU6?EQ5;YgDMt|fXuq-CUF)x{=R-1e;H>4=fl{Ur#XEKEF3?=M6Pp3E
zP9(eu5(6NzlGKfaZ
z-!|vhG)joF0EF@r*!+JIE~UG6Fs9OXq4=UMe~$My?2W7)E1M0l+3~b^GbtYgZVdCR
z`oHngvaCFhNQ7MCi~H4o)oY6281RVYPA-HVWjXHOhHHS<*zh!cHc%)ZWMC@CmO_M4
zPH^NO56ITP-Y{#T%3%FSot0+Qof6N68uQk<=y(o
zzu*VPXlgh&?F>KR`4uYS|A$Nd|KNm^E0y#X486gZT-h9y=l^uuWo12^2JE3B`SN|O>l9qO{+8wlw
z;P!qwMWM}O*SPCzE*xb&7lqbmJ(nu|!d%Q`E65m;FkzPX7%~PjLU3$h$)B*Mc1Hap
z;&c%8tWeCeEX=h%Nr`al;D`j1BAh>520|*GL5D^KsQUf`Qk2r7eBt1>%V%hLCA2__mF
z0oKn5tO9UpTpaBnJp%ekP$-T;D;!%a)1Jdjlogg|qspOhf+u)>g~pvoFp
z8mdsg=|(GPsnvkFPO6{o_5UDOzjLvS%`i2b@B)74WkphhluCl5Xlnd?z1M`ayszF)
zh9W7s>-s90YTwUw)KgC$>m0-w}SUTnmPcmp5KkoI-uXy(Md74K753e
z0pGGW7IGl0N0P$$&^x&#EHm9$Rol9+L=W~JA91zMlH6aTR*y^`0pKVvhGr881YeOU
zD$4PoTKH+}++cPHw$WRJgop&$1R!s~>J|_`FTG
zo^-0^VAi(4&ssexozK>Nlm&|V4`XJ}T
zUj<{djjY$f7GiYp-q*2+*?%kc4Yz*F7Uc|f!i=0m-!>d9%`BX+^MIQPm9Do2z6&2!Ug601YggvZcT+kT
zAt(Y!9wEqywW;KH$%0|h-SEWA@W&Py6S8hR!+%$14D`gz9?bVU;MtMc5ekgpGCRP4
z=v(i_A;`iM8YdHApP?VEWvCQu}a|rrB01Wf>V7Lo`BQ
z2Lek{!T(Ex4}J?-TmaS)Jlfi$xLV*eU+wwd=VnHB_J>c~{@?orNFQV=d*5U$8zK`M
zn3k1BSC*zvHPv6^jHi}2f)7EQ#XozV317zx-;N2iJCQHmi%hl?nN*WT&z
zwAVlg2|#RZ0S3@*pBo|XZTxR3h5!RVTvA(9^7k7sbjU=df=?D-*BV0;5RrmtWaQwm
z12ugk2rge)34%QItilP^_QgRC;?0X6zwZOnvotyjZv<8!{@svZTM6RbD^v`i@&y>Y
zIP7`Wd~i8|ok)C(T}*&@k9MC5L_KT|K;@;rW^TEReJNaPAAE}uAb@y>c1HlP!ZiJbBzbOL4h0aDcCU^zUz)evFyC!0Ymjum8z+-#3v_
zs1Vn)AvSvdHu@(})7Y)GI!j~_HFL|{&?@M`PK+{
zxQ4$ao)r}((L~ny^Gs7{wzmVpT;HX%<5RK%bbiq-oftZ#cH1{T|9y7&a`^`yAgq~(
z`Z32p@|$nwM-fZQgl?lzIYDG{(p3vsl#&=Nk~588IpKAZ^c5CMeD_SvDpznN3OQNt
z<09Yabg@=2JIef-TNOo}^WB=3iqU$H?h-z-jQWjKEM9gw5Oz`E;6=r9n6IwbWez(I
zxJugS(UA&24m%i6JynNO7e{a{jGO+-zFRW%Qf8EBh*rRP`e|7mMcqQ^8Kp)@Ot4MF
z{ObpzH)Lw5`NMs1oxilm`jczEC7NEsKmrPbMo!6tPd1m2yW-AkLWHNLG2Zh-@?0**
z-nsG5C0bvV@PzCYY1L6}9ouEcyIp@hzzSw*ETJyhq$bEBZgV<*@6@ALc)=fB)#fSy
z^&VovJ_HDSN<}_Y#nW>a)K$&Ztrnas6B&f37F5yQ$V2%z=wr{T?lqqti0=x%%m;ir
zi$vDN`eOMayyOTUc@bIe43PSE>i3cv+tm?+qh?}0MsihCmRpe;hgv-{Q>7OJMyjfZ
zBK!d^9SpOBvOEQHaxY{7N8|S~@_G6P5rIxcOy}Ijq9O3B$W6rh^fqLS4av07#)ZRm
z6D2rR!N~WB&LW%LJku@mFdi9JA}X@lKl;mKL~Zy|ifavJMP@Jz)=6&#dyj9+Mh&+X
znmo|-rp?yiAP4UV
zqTC>sLDxRyu{eU`ridguwEEtpis*!|uAHO7h^ug?Ae6fm{CYp7A
z*$HktdsQI0R7EQ%yL&<59X#vG2ZzZ({m!9>qyx>VdB=sk5}wNWnlu{m1u+h*=h4&M
zj|iI=%xw@RWWaV+apkiGyo{O^&}yIAR-<#x>8x-fgk;!LTC;u@&6h8^
z@6UXEkX0s)?qsynSTiSa5|Y7Ot74fB%Y=PFTNUX?05-=p(U@jsqY
zLRkN+yRptvqK(Ds*f#c%{y#Vk$y`O|qcR7wlhK6{Ezxow*lB3oH!Ff~*r3}ygj`f-
z$<82;L$L9dLY>o{i*QtIbC{0R-oG4;4av;59paY<(YBxfbiSL8NLg$K7stDkI=q0O
z5F_$zMx9E_Xe%TpzHlTy0gw~H8)!(3H_McL80pX1I&mQlBgqIx5
z5V1t&lkEX7a{jiJ=O1}`e8cwvW3p2EXGk4N6T*(~59?btR9x2m{jmysVW^=h_fK?U
zGQA~4)Z@B9tdnr&+K=BaAvFcW2O;!%&jRW6Yt1P3{J#5RbCvw64&M7c_GW9$A2>^U
z?<1`fFi}0`R&Em0HDi>^AkcnCu5(wGS8?y=hVcnMkhjt5i=1KtC{?#7vTrxo1QL0v
zE7;IbP(GR7omty-4;J43`lB%A#OdHVT}#-s5!{9W1vjuP2uG3&LVkL5>6j*B`Iz5W
z>!OCz^jRaC&DVB|*lu%@>l>$g8Aev`Ei#%7n8TO0Dcft$rRDQ&zOVaDG{omTA_TIb
zlqhPY*_NTUW_-4L7JWatr^L-GZtoL3zvq;3nsy|KM%66fcG0>ssa~!@EHx#iC}0bU
zOP5mt9R|)>S;gklbLWd*roV1$BBT6ERHo`whR6~6H8&>JfEF-4r3=D
z3&I}uIH5`|Us%yhi`)+ReL#7)q{Qh!s&ZovycP76F_xtMMog4y${kY3E<}DV4W&en
z??l&ObY92UHC&j?RK&?4R*~L0mO|Rmy0Le6DM@1U8}E`@dJ|vtrJm#*LYVR
zS$Po0#%aH0P57PBW{HfMTO=r`b-;yFn7_(l7KplI(Gc~KAllKY!=)s9*{(0nPLB#rWAtZ|rxM|%?S
zm+7HbxP=2E(vLn!DelQ{11)|idf?yskHqoTxRW%N51hz&hB5G(Iuakv$^y3ybOglg@TU)hRzFl
z1VyHL{SsQZw`jqDZ-Rw};0O*@TV$jCdXiK+%+DmFM5I
zFqmGg_9cygl7K#k0r7}J@nD{`yb#Lp0}8xQk2>6FF|LYss<+j622M99&fjiG@P(5D_lphihFNZDs7
zkMproBx=xCqN|<+0*|`FfPZ%t)~ZXUQsW8oTaj~Hy(Q8Hauus{bgVmh>dnp53MI1K
z;1tg@3>ZuZ-87~odK!lB)9??X6c<9g#W62W_@=DYXr+w~k-R5ev`g6^;;{^LOG-6E
zrKu83XpWmAn2m37(GX;~PDTO}=@zV{O=DV~8RiRIsZrjI6x|8W0ExN2Ju(X5phx?%
z8-uGWWly2q%2|RDNXo0u>Q?w~W);UxqP4ZfnuzROJ!1PHHta=YyA)?Cz~^&lqp9>{
z%VIQoDYqMAs#xP?vEI}S@+HFD8GG1@55v*CYS#p#=k=aMzwa@`7yI7L8qc*Y6~SOV
z|5mMX{m|*RrJ@`K;Lfdh!m3xe
z=-ptq6WfcGTA9B4YURIkcMY<;%-zKr@?dDN!@DpGX|An!NPswytME3Md772Hly`B%
z4aQwGegy7~0O!ZyGT2gAvxF?$GKe&*TVKlklnbp8m7#(M1V*lI;lSa;FsL2KTukZx
z^*%|1gQky<*R5sPT=dGVCQajA3vrU$`FPrla(@=-dq8>X~?xNwpzQ%VRg
z&urW^PLXQKq)KuktX5$$6z7XmT9?b6xO>L|N2?}*5oWYy9V0Tt`+^Q@Z_
zMn}I(#TjE}hiP=rTxIDYRK1I2AN>A!DR7_W&sa?3T3&+v3;s=F-A_iYYo0U8{oKNS
zOIHaHSrFr}nk;4_?BO0?Zs=$HeXSi>q0WG|r>Upsrf9&P5R>mB-95QPn?D<6h11iv
zx_eSQ_5vPkXwI5jf)p81L`;eVh5b@wLQXPbH;X{ynyg3T-b_+WO`FtsmaZo7J4p))
z2jNYtcOVyJJzp?#4v&T@wb0y9XtKSiNSYXM4lOz~s!xKS=Qk%MHdf+^7T$shydYou
zCA-u0H!A)wFEz!AW~Dz8q;LlUeWsTnoX0mpJsO!TZn=(FI3fFpSz(}3fI8AspRVM>
z+8yr>hCNm-AgdO4qc8(gpCDHeBNZ
z-$knmiN6$mx{VIeBR)CKwj_BFmc%5zY~74w95nrOIYk`M$U@bC|3nfQ_b)&#C|d~l
zfyM~SuCJ&%X#)<~GJuyYQq|
zSDHvX!?o;Dd@tVgd#=oC^md8Yb=C&xj9>zpW0!AIkkVY8{tJs59c4*`I~3Mm1_}92
zG!ydVX{F9XM#lFx$lf)rp5=|us7WrXUV^?Zb4u#9cyIgH_^4W=k%EC9k6ZO@p#*siMUD-Kkjgb)
zNNUcpzIlJZDw48QycZ2oQWBq%$#2qI`6+3M4
za9vS%wbHwbsX0J`UHeM_>wdI$dgFZB`^7Yq@eaf@`zSQ2np#U>rB92UzOr+UfFixU
z-=xyPt82|csves9W|TxlH`EUJH$J_D*V>5E_r9bo`(@A?y|+{U*4DU_tDbFw3z$4b
z8(?@>6v=&{$p21B*T`cQA*LbIcRL}Tb#mdp{4sEwPItiP;F5RNv`2CWQ|P^3Hxe5U~K|u78mFMk2u3`oK<{fU>k?f}mexETCMt*~DYS<3ICgx%lXKDLk
zJ&q_k`0UWd_eH{*6D_ZYYMdIm+MfvHz!rcTjgJecin>?&Pk`#dd5a`&BX`tC;I;lu
z{PbZl)z_zYf?YXl?Z|P+nE~&uRad*>jJ`ZX)h#`ekVbPq5*Xm<8mPBF@fW1~tX_fq
z)NTC_Y1LR*3+&0s^TV_KLi$gpJC`UQN@c
zxc06I9<8EMAuSW0zWKh1H>mDa_eN1xa@4X+pwcC3jxchAc%!%*W-{3Ky{I)jG8@O}
zj}9OAogm11`Q)jsMPBfxK*Q=CS;la7Z9HV9p;F#;_94dNt03T**J#`;B31Kq>eb-z
zJl8vSf8N!quhhWf*J;<9Z~s1FjW?U#%Rku*r^Y-UxpXoik?HWwbNavGejBf@GvGdLH+a_$_)U05|31$kd!>Mp6jZyltChhZly3Oo3N3C1kjr347K(S}Q}ezHM1
z4zW_mKT98=ld5{$`B3zi5A(!ICWu>)C;MqK4^)lAS??)F-F5OjXsonGHHNEA1kJH1
z2s&9mwvx9F3uqUcbiECIsCr}FS}NVqM}YhIF4bVwL`ebsHW}MykXu=xM!9TI)y736_>kM#au_KNwTJEKiH+65cbjMno6ud@|SG
zQHqWKev1d(!~}hW4YiF2MP4PyaJM1HlbMA`v4CQ(Ks|eJBiHibn-kJBtDJTzX9ibn
zlg3vM>?(%mCP4GGv*0raPw_0t|I4(ojn*pd2yH#Jd9pFxbqj6tti__u!O$>8Zo*+U
zWN>Jc71etOa&i#TUgkb{m~f~2@`n6@_VeWaVeKQ4yfWI&N4~h?Mu*PwR08Y54acH=;On&*u&AgF
z%|{p+iC&mwW2;j->9$$lqYC|l_QQxU(clMoIeG)9l6prAv(g8xl9-s!Q7=6Pmlw11
z^*SJPt4LdU9rWzR;l`nVBnO9F7DyJCmX-`<{;x^q@dwALx8c5QUtW^;m3lC-dPCj#
zS?zT25!>ibRBi>`$5-}ZN#KQrHAS60jQx~6+jPg4iU&ij*2*U4GMzlER4J>b%<@Ju
zE9^8(wYhc8=$YAr){>+djlugX;LT+rI+1*?84*|uGP84^|E&j}fp6P3|3PKmyJrO;
z8_QL9y&|iq17>ja@gQt6WC^Q0&tqqm`8=6YA!ezQjrmk@p7r@>cBo8_+~#$8?sN3D
zSgvh!PgY(n$?ZTX(|&Vq7p4fF`}Jr)^J+xcGv#UHP6K$t;NEi6W#
zDKqXy`hEO^e99jRk=9-&2EFIK!#@EK^!?rB1Qz_u2Q!(GIQ!H_7Op(!?$e;k$u1!#
zT<+7@nX5v`7A<8cM9ySivzHxyKLdvJhvZ=gW;2q0?C_HkdFO_Amj5b9ayvpq>x2ej
z!moE|z*Egcft-9;B|##j$n>!jPaSQH|F>CzaCytBp_yVgMlv>XH`f`N=lmM*oQ_hi
ztt(|_>`IiVLTtDJw7#p$?Y!VzIl_vaSz+P$@Tm;{HykR@(?RbNW7OCJN={xzck?`Z
zO>xllF2O9&%rV!88eGbML1OU_)P^KFDri>G`U(H}&P%a2>p~i1QFx0EHMHk9wdB$H{K@^SqX9L2ilcoLHfc)>fFh4knHkw`55R
zdqY{0LR-~OVFH!lkt6q5sD81w>v-rKz6J*nuFAuGV@=A3@|Pm|=R=xCl*F>Wkk?f_
zbT&2XCyRcbhN=xJ9I}6VI)vMJgDktph0?;%SzHp~%~mKk6Q`}>bA17i#7X#M8XyZN}lRg8$ucwLu0wVS$N6>a6eDfJm2~{{Z2pamS7OX
zy*b}<3^WT(p(4@SR3JPYpIh{QM3SFa-*s&A-2S+X3C7NH9;XN(EDV^K{XYOJK-9k!
z5_HKKPbFNX)>=}`ckLg=xy=tgQGcU;D*9}giQktGEXSKfzvbZ@)+85_LD_9*wAO1S
zE7!IqcoaeAY+`-`WHEDu1S7Xeyk8k$)svY@jWS
zc|gOOad|k?Y#d&A#U69Tz!X`|jU-{09LVF)CN=TpK?d)D;mJsHax!??(|{u)zW5;f
z^%tOx(PW)RIqX5dmyC*x7xk)VNZJ)Q($#aus2~+8eD>tCpR6L&hF5H3;;r;y`rJn*mLt02*z(D5YL~rwOM&02so@xAU>C)LR^ng
zX{IZ3Y3apqwv|qW9~lGlp3pRMPT*{k_kks0v|nGOf#WTk1Vel}Z+~%l*hx_EIyf_&
zny6T&pN*ZdObb*zZ7+Xz`Xzkhk&GPBPhXR0v)8DqcT(^~P+Fp7Wf8E~d)(i^E~K(i
zMbuQ{ZS-NOYe&^FTAV%~ip;I?5rg?S^GdEn7rQB7c0~X|Vq4m$nMsShUmCy(wRO0O
zZ;2I^AHC48x$X2)3xCx-ky5*ZRqC734R?g?$}7@1mN8O9yTe75=Rk%K)q>(@MAttd
zT*E`Im^1I3HlI~PQ|2R#e#_NoP_V&V~k?h!kU#W;!%+IaDGG9nJdcgSHckyyc32NjMpMRe1FvqRiskwh2G9P|I
zHq3P*lUN;6GQ4mOqmDqrb$O@?x%8)lxeTWUdoKY95o6hpSVgJ>$Yx|tNc3a^<`vbn17Gax>**F;w3zi
z2lRy7`n74u=YJv7Ar4!@lbB@^l40?iN)-rgK$9FgkUT0b*u%Jly19P|wV{?tNqR#-
z4rXbdtrKK>d0=}0J^S%*S{L1AAX78-JE}J?cDiu{Tg3L|Fi9knVHgNI)G;u@2Nu0;
zUhi{NFS%ljIHOW$?R}(xJ6yk5=p|3s2aV|njjn8JPk*8PlGwFE5Lq86<ZlnN6!pMD?JbKiJ^vJm?hKh$lSDP>}GWD|*jPTv{^U
zV~Z+PyF5na&rAcZ;LP1EaTrpzk!K@b$fxipA&PyGd&F4>huVunEcZrI;)5+4iG%Nk
zFTf{t{eLfovFK-a4yrIyD|%*^iw%JCl8uWwh|lQ!u|{W1qpv3bKqi7Ba4WJ}J$$gBz8av|U9OQ*OAEHWcs
zb%^Gpf*11=o)h|A6l2kHS#Mn@9!zW9a(@NCj~$zUPZL$xZtCX2?HuyE-|&^1
z_(#V28>yn*bj1V_{ZwssYd5Gn1{$^pA3X(ndVG$4ovif`xM^%j8y>Efx?i0Dq%%3x
zrpOAU(#^48JU_-?M}4aSvvE;0pewDpeMdE_t+Hk?EE|lStXsidrW~}U;*IL@njZwH
z+JE^Yh~%|K?i(=}+Xga&zBu&`xp?hcVfjgubG_5w=2@40Ua=d$5#~MATgFpKxdSNe
zhbkjyl>3n|+Aho2U>))Hr$FgAd;0o%|3}v
zUou`lS>=za<91F!8Jz^~u|zL=Gk
zuX(Ks8v_=|i)$*Z@$o%xJI`(5%oZjcx9T*kB5N%M$VFn6kdHQ;n;$D~AeT`X8h^?h
z_^q>NE4B49{j}wSI~Bl?>xki;twrW!IIXw{#;!@i^7Wsj*XgxeR~E!zk@}Z?yExOg
zefUuT{0b&pi%54v!Iot!xLa|aQE#{uUIlU;QUgA!9bYC3f3-BT+%Aw*DZOF>WV$X2K9l!G1E-j~y
z!=UmAnV*9=xoTQU;*)V8+OlzvisD;aX&bNRjeH}i8|yl?4WF8}I_+1Sk5;*@IjBD9
zIqD{&Ya(QmlBTl6`rj&lRDZxsdfg0gh&W?QrS`B9c0dfBr|^?4c_?T?ebw_upWi|@
zs|F_@0#uJa^nOhCMmoaxWDt>rl|p?z;rgc39|P79pF?u<4P3rNVBo;Ci%xDq$+?S0
zKI5>%rR1v?p~rD*4D>A>l4Rc0E24O1w5#Mm-x*$SxO6g$1@knA>&SJBYw>o3$7`f2#0;kZSNW5>g4a_+!s*Rzd5XMgLiUNdpI@6=e5dTLE)
z#&|~i`B?*c?a6bt)_-@1anK2-1E>yc(iX%t>$~TZqTBL(rXck&I*-7^i@?va^hEHq
zk^c51KAR`fB1goBRs%nSuA?wk#}2WHwudpXSNuk`GF2HYPON)BqUF6v8K=-D5)K3b
zJdxlJVf`&M65}YB_z9u%A%j>8OprJUyGnNbs}&AlZInl%n16DHuqp>p(D@cVGU9wf
zaP5U5@Py5M{Pm{954pm%XOeQG)7N3r%+vermv4p%U^u*|2EpTTjPUV=W9DQTKE%pZ
zRUvcEG3?w9b}1>;{dX3iZgBdhW`$>EJbMbNDrl81&NF(BA4jV9ny>4Q0cGD%NhSBl
z@s@`|msV*$B#>yjH>MPWt%d`GREc#&W3A$$tSAPUd3-KYkdA?HUXh
zcCY1^!gRH0e)9Fu(|I}3)l|n~OpQi~VP?(%jV^Nv5MT?7S~8aZ**KG05@P{c2IctP!vGCy3IW#vTXcU?Wq`2FAr
z$9&P3CF(hdh^R2ig?7+_D5nW#E*oBMlyW+4oqjkEBT{Ad8`xH=U+C{K4TA_~JJ0Sk
zkd=On@P8_YKVm@?LPhZG5*iv85S(a@pO1fO5nCh-z0B>J8zvn1KFdc|`K>;ts&?9p
zJ`ZPs{>xpv$QBw#Ej|yXp}{F?wP&MWM2&4F{(72M5w0bOyf5d7Wy#nmD0WM$7%3*G
znWGWTi->tsVWeiayNz!j-6X`OeNc+qY(fvYgMWyJeu{Sx44+BA$rnO=IA@i*w3?!p
zo;%Zb3S#5iiPg34L>ZFVw$F&LjwUG*T&_K(ZG%3XXKB5T>SjkU`ip9mI%{)$-fdWa|@o76M{*k~00c
zk$=4?h_x%MO;T20r31%@h>nt-#jB{Wpk(q)6$HMYR&d>OTW^wf7dc()nyyVVUDP25Du&~^wJ+WL=jN%0~A
zs9HC7Y-*4~b}`|7t#{G
zYZO%l<48;?4*&BI;RUrnwCZ#4421$b-tR0%3I&-)`aWH1~#3lHNmtm;|MKyp8U!avp#jYes;nZ7k=o9&3wUu=Jr
zL!Oa{F@SDoX=_LRw3YmB3y5r57CV0$QgWdqg~EQn($jbpoAkVmBjbsqc6Z!5i!
zarU~CH2tfiEPcDMIc+Dm(xft|sb2?$1oysDpdcMqxm6vvDYA(iZP%iBZkR-iJB7Pm
z-j=kthyfm~8HvzBqDV&A@3SXgma=hY_eV<`P>UH3Fv1mkIAuJUP}L@tj(=16M6^7S}zci+_B0Tb?NYp)IPPn9tuo^p?S0FA@}NA#FGAW<8>!uQ!$;h{*3TR5AFUJrku1}4Y0U-Q-aJ@K?sdsJrJ@D
zgHL~w>%yME13z9@1D24fB^-ooD-7j!^XGfa$+WQp;+I85KkP~X7=K1Ul4uP~495bc
zr!MWMDvp#%;4nur_6^Iic}5Kl^EtRW(qgs20V%^m!fD^`Sw3uXo!IlQZ(&(sOA?(d
zJ}sQ-%5Ue;I&F#tc^0Di9+afxl6j7h>D1bsit!FA!SV$;Lo3wvslH}pP`0(9;vARa
z{zPB+qDJF_T9{s5+ke7+zs;~y-!&45MzlvZsNm4<`J=^QNnH@8|G>cZm&nDIpyT)>2;WR$#OXb2_RNALL&eDT=^Z*lOF@Faww+_B&Am3#CuYi1X&idis3I}G>@Q*f*V1{>^k+T`Gui*q!p
zP;0Jv{tFs}-$Xsro`Z#XdtUS2bFuoEIbGKlV90I6t9WL)`K}vefh{l-S7xgL*86&rEH`=d;eXGje&QSy|i2$t|#xFqL#Sj>p_qIkIid3hYaH&ac)8RvBJ3?OM-Ezw@?R$pKWvCZ6RViMII}h
z=k#XQI}Pyt7O1YA)~FC6F3!}dkMYbSy8>iZmEe}f@`&p`kyw4@dAbT#iN{k
z#WossI;5Y5b&Hv9L|~~qMw&cT>u2|TO_o~O8-JDC*Ku-U*R`30J_xX39Kx)Peh9#x
z{y}0wksZvq4v@4m!DSg0AH_`=sy?u$74QTL^NZ&9e-t_t=jx=-{&($l!nviGxJ#|IwTYp%fYE_{Pg4}q-iH#qI4Y=JF=987ueDVms
z;c(>QTLeVBvCt2U4jMI$5sr2LiLZO_e(A>Iv46({;RS9h>)qX*ttg4CZ3+<%-0
z290#8C`Cv==7mF(`)f4sRCc{kRcYzv1q-=wnq9L{_^WOw2E4NI84-vpM&OyE^d8J`
z+$yu|*t^buzz_PW*fasD^e8s|+}Q`w93mAL!PnIBinL4weUy=~5I^uJ(gYj&nUl2D
z;TX<4B@&m(rEHtq4CRny0Smh+J%4X-;wx7EqrZi(6{0IaW9~DWw!83%Ee4Lwr65q+
zZ}OwLX17b%eV@nfaYN3h-Ewd+WKLCl-igOg}rc(~<-XHc*!m42!F&$V|;^UN^1Apbrv&-`I&&1P7>ZF)p{ebw
z*15@Tvpe9Id6n|Ff<>TtgpabWg`#r#M63zbJ%p_rAG_KD_7HGBGUNHT3VN#(
zrC^fZ{~iHkM&J`q@hPd<{*oEyU*NKC`WbmFGG2P%eJ2{<_>xEf)YIzr@|;m#D$X(-
z@#SpKGjD8H@WhZ)ihC5m-gDg9zHJ7hh`k^46@Na!gcL+wWK6-AF5t_VhAwIGHW%TE
zqX0uB8}=gr&&^G|r_)uU5sMV1G~g6`jAl
z3QG#HQvej_4SxVX47qo5TUo$DL!nrgl7+^ynE%KnspPBK;UtF1cIrUPJ&a6rF!Cl-
zg@0I4x$F!}d+<@Xt84aa!BH6H8BmYlE9IxoYDqax;vut7MjpDUU&J{A*d>^^`nQi!
z1zFsm*L%8Hb0o;2JwFy1S`2=mr-*-(YSN`K^dz+;B<41+g1BHEgYv^t*l=yzG#&3b
z{GD^^ud57~xI{Ev=W&`6p61VG5i|sc?SJH>f1zj}xVnr0^{J1`8X;&tc~hSqtG6Z$
zvU<{SGc!8LDCa2l=Xsq2VCfk`VnKgdWLx6AJwDu`5$0IZQ97Qj-Mb
z_FO)4lfjTUd8_*%L@BepwR{%Ti+_v_&`UqIB>pyW9(Kai_Db5HrtQkxKDeLr$WA#Z
zoS5LdsM0cCX^tG`T&Yr&}7HaRCw5VHE8_R4Y6i
zznl13($}<&dtva}WN{GcF@Lir$3d%@ySXX$@#YCga*S#jE%WerFv_qDQk~gq9c~#G
z2|*!Ihd(F!6Wo$N8|Bw`sSkur@b1-abyQ7pc3N|2erQ=YXAFsL{y8)hJ~ib#Y$nxk
zl9tj+j1=7sAJPn`OYY+^oucxjwrn(aT2}t-K0&wXeb&f#w4nh%Db(fL6WopC
zy+SS@Jc(y6NYT&*fgwUxqxy#jx)I{Q-=ZB9i^z*JD+`|AgMSzsnqz^AbTO75o4VT{
zwnFCPJzKsps$P8V6XR9Nxd)>mW?=QzlP58cY`u^9j(qY#v0sD3N5*5qgU*m(&xFh%
z!j&$lrr`rwu21HX1~XAIj=*u?Va90!U^0@gzpI>daAn?ppf&LBg3gN+GGc57=q>)0
zP<`^Wx`lkZ9S7$d>+#uw
z;*Qp#>nOj~g-_nOPMgwf2}!Pxbc7H_w6mg?SX&p{OxMz7TGft4dz+4VUc@5LS+JbRV>_OK3B)dSyZt%
z5UHtV&3`g`j>$I1{cDe}et_+3{w37srw1%OrLf4+Z6=D%y)6haB~>nnO%@c=`WwP;
zg~%rHBA=1Ud_#%P-v}5`iR%?xZ+|4-5-f$8Zttp_PY;DkKcqDu(v!i3
zZ^GLlr&)o2t-C3^dXiXc{~1Mri0Lq80vMh`YctU}TbQG&;YuO1rYkC$|H{bQDk+vl
zjgwozb>K<90^WU-ZH@83;54(Fx|5l(IBq8+B`dk)jO7tIC2-x10S82~{qV8sC)0_7
zjeo+P{2Am)uO1kfLWwfOaA)69Xu#_Vq##UbL5yg4$t}zq^+j^`g@dw|y<4KkKxLGD
z4}^wOpD*}VF(mJmm%b%ktgE_DwC2mwl}3ruA1Wh5>QJ(vo!Ree^p~k8fElBCTg4*u
z;&L=!q+mP!8GT$g(?en`qi-<3zFAK6(tjMaq|G@=U%`xp7amCLB$xddGD0;)iMkX;*ZvviE
zEX%v${+bNMaxY^%#y+>-Ee@4U&42eaZWyZp>x3ujJ0*cdiD1T}Uo+|%L=#v!8gfeY
zTf=_P=)2~nL323sB$4mNkn#l3EtHb$m|?V%?;T@b@x1o;Z7er;o5U@{*Cp8iRTy7cVEhETVm0RgLZU*E%`+i5?M)Po(EH=-^0Kd%G{smyM=|hku@y1kZH1
z@lhx|n?lZRvV-eOsBltiS~fv!n934tGzMP!twLK7iydJO}+YGkkDYMl*>qpyDnES!RF6MUYzgR7Bo$+b&4grGNOFFiVoMb&?Vw(%Dbe&FPq(OP8aeN_>EKb%7RDWS6(f
ziq;btQU$rtB#SBbR4Zz^X9|JfH(9*Xfiz|JiBcc>$QG)IOakA4AiboD<-%)7z>$`t
z^dbbxp6jdRl;kyl3yij2os484uYdRQox&@EP*+1e4Gdu)fI-*Ri(kSk}D;Mw?2ta73
z%-w%u4jxV}a-)8!gb0N1?PgOso~opzHfb1xYh#4rVI_^$h5eVMu+
zkcGFw4tpgYP|KjLp?|AAReYAWkXdJ?U!RS39&oT%e)>waGlI$VJ>?|UW8m?re4A*(
zVaVf!T+foEyS0$+&CH>1-MPsl2vH@ulpidD=!|4EdcTrH>*5BnENGhOOe5vj(3uQ{
zA!A`F-mMzlM2LnB#7L(6vAxaLufdmQ>*-D>+OGY{Lc6#a_@skD8Cd*iQqq%0<-e7pzm%~)26xAA
zt!J=an^#1s$9$Z>@V*!T6Y#MXp^HDG18ZJ$pPDU+<&SIU41z&gN>?)>jf>06!*Ea0
z-2Zh=rJSv84u6B71fG#02kN=oWJwfG8m@xN_Cs38=2r}op1`)5`NcR$=;=^
zg+v!!be4eV)O0$Dl}P{vQD#e*@CngcwdKwcn9Cc9SAX*g&P9;PuVCvK_IYSwn%%km
zH8y;O5YD*Xc&#v)WMR%Ki2})b`Ac<_3r^~J2%W&%Pe(7g!!2}bIQ{3BJWX@>qz1o$
zFl1?-(-qwLd8bP;vF;`8b_&}p4&zaNQ@$zp;v1W8C(&g+PQs;#QD6uTzehdN|B>#eQQ%w9540oIS1fmhQ+IPxB=}EqOeb)ooQO7oCvE-r(L6>St{@8kD;ch|Csm
zVSjD+!q6e^Hxw8ZkMfaqz^b29HqT$x4^~FKCELgy6&qsk2K$7y6bE$n8}={U(D&84
zsuLZ&aG`35yp-iH(ft@pv7JIEqgWW-3{T}+$T-2h-O1M;y8#p6WMVhhjvv6rCX@%s
z8Y4eQAs~QzOYo<6poS*WtmHP(Q1{>j41W@yGmH6+5$ENFao_QzjIUBw`o|sT(6Jw!
z1Z3?*6rY2Mk~l5tdNpsB4_QJ?@2v_z+;dJE3h#K1QV5;yWMNzKh8WLqvsGq(i*WIx
zm$f)Dtu_AmN~XCu>~lR}3ctp(i6s=sL@+Yp`h-6uOr>dhef|I1_QvK+7>3u{wtsEg
zw%e=P+VOfpF(Vc=OQxC7<7kf_{_Sf?&{S&|od
zQ%~o-=EP8oSNyKgFX0Uh*8gh;Fc|wtyGWME+0dJt3qi^{`bKy8sA1
z2UvGm0d?4Cq4TGePq#mNDZ{Ge9pMDVG$ypACQ752Qa5!eQ?8
zsJP>w1_P=%&0*P7RN0VrxFp4QDA88}3Cu$mAOKfk{71`O-&C*pgmRmn(#D1lMvW;O
z_ND^3TJeJv1G*a1dF2vBL-h#JOOTBCIXC^Kypi%l6GG8D2y~S>_`7h7g0;n^jQVYTD
z0Ri<)g(J0t;}F}?Rl7%4%7)pW$G6C|5^y#@l?{o@#IZ)hwB?WY6dCKV{xQLy@kT$=<
zV0QYyZ1+oE<%>oXsFq14aGpc%KyDlFMYE^lz$eW3@#0Vg_VfM}WAV7a^R;v2ob>sP
znj(Yo)vels8QPwZtwmKQbnQT8Q0dR#b6$}$%4s^}&?nnmW>o5KN`EtAWFYS3^2vpZ
z^o+~%tg?*6J;yLgH^iSZs)9Kz@u$X@;+GT7DJC=&o&r(N
z{1-=7Xo{ONAqNf0n16`b4H)AT#k7Ylx1M9Y#wPQPF@7MEF2~G#WJ}sSYd5rIesj%s
zlmQwQI%b9Y{;%|icw9Oc3Sl;-q((8NBbN6@f>xiI2?gxWekk0@w3&pA^Du04r=C{P
zvksZ$`%E~nQiKcUYq56LiL4HyL5}a0_QD$1Xgl9fIlp21Pk-wR@bWd&^EO~8>yLDk
zz_F@X$95(f|SSu3NbevDTP>HT5Wr
zN)aaIz?Fe}?RRaabSud-Zq<3)XcrX7qV#Y=w=~MLb)=!Ug
zB;%2A;q(^^@P900!z%Z*q*ohq-%&`cV?Yy%o5nm(XIPBO7?$kL#ECL4lu$E3=4_Pi
zH!l>i7TIU$Kv`>ty$77n?JKsVe#NN7Wed)Bhf~&bt-x~Yi&4(^v0t0rH{F>G_)jUf
z!${}ON;LIZB?#zW>Z*?;7Km9!Iuh{vYP&Hql66*^&VK;n&{+kIt8#pyN-xy9s}FaW
zcX3hgdNp;0UALCF8%h0SKNS&QDERGC
zw%Y|^BaHrh_~bYuyLqn7e*kT>)_;66$>pZ8pr9ODE9=5F($Ahp#&R@244
zBE(=#4K*EREC{{4$w$54NgspJ+c}t&{x>+WZ+`;Og~-cLlDG*e!yamyFmK9|cJ?EY
zLYp20{XhZ+L_gXIXl@5Y6bCn|H@7@e#7DxL{>sKzLpvt7QX;s8U=W27bI
zamC}Z;Im)#-hAS59zGkl&4^?A?PV7q*O>iEGmN%kd*C9%VMKhAEiVhY(u=bF17VhY
z4}X+3SYN3kXy+i>Rhd^d!gkD$*7J~eKtAt0G@#WSckkPMj40ThDhZRL{^M9
z4vZpfUz~FA%|gSejDmcJ68;218^ttCY=pNteVrUr#5n8(Nj#Q2fy2Y7FCP)(`%xzX
z-aAEX;CB~WTI#3?+9orrNKL;I8=$wbP=9_BYbhTb>zSjsgxPk_V%3b@xIdB%sYvgb
z_Li{xkeGrO!a|T5c#{uT$4&qEDO08a%nTk&cynpk?lJ~FVOBQzQec1KT_B+YW+uEg
zM@GXJC(bNX4&R-UYGxsEc{b)uC76`3NdK3%V@07gRoo<=I=^oShrh)IWrA)qe1G%n
zH2$x<3&{)BCoBl5!H^>Sy6u)4)wjjv^fcM^?^D;Gy)cL;${j3QWy~c1ZSjEjd^H$<
zi>U}4wO*hFT4oVR*_izl*7*voPEtV59jQCQya+y7D%au8R?hrbvo=}LO!4EdRA|B+
z)IQv&
zp-|PE%faK(U6E2>u-_!l)LW^-Z-k7k567RIU_3nq$6)3My0^B$>le;Rn=%b-qBilYD}Tqp*)T3O
z59e(Y5!$z7te<4k>lf_aj9s1bPkT0KVoq(s2?XhC_?N<&zKd;}&gMEdA#pPx1j0Zq
zdl>2WiFFLV=-FZ9U%QWuV)I^R^VL=&)8hDyxEo>d?cEVdJs)Pfe~bHxW0NJ>X1Sj|
z&qBQG%G05_7ktVJ3b^F=`+qr?EKy5h6T>lsx!8t4&_41QH$fq<9*%1smd6OxX}B*QXNkW>-YN+ZeK~E>OiQL7~?i5Fq+`XI8itnlF)+Rl)F>u(W
zebeAD5bSn_Q5Y9Kps(lx};rFPGU#Qa|B$6QzG-!G+m8c0cnsM!>_e6
z;3np_FVt4iOSU}?rGEfr^jBQ|w}>(fXku8gQ!J#GPyFL;lyTx2jH=%Q0_`TGMY&`r
zY=kgTk0gK)70VLyr8A0_{5f@dwRfj*;*p2=%S3
zZ4`uFzminc6ar3fGg=M=iE!Yj-~#3kFGTQdfC-i50wQ0}e}83;0p(G_^!#`&9?yo|
zR|;$(b#8}B3jKlgYo>get7Y_kS@Tp+!Fk!fOrW1s2xryQLEpuOOi;BPMd<>OGRh_(
z;OxoW8-fU=CU$5vUCsCDH0UaW0@sA$x$;peDkG|CIPxDozx!B`LaKgL?{v+7!39c^
z1OK!zX&wvaY=4!6SjPcS8hR;Zs)6d-8VGa$fDQooqgxC=tDy!v4vFo{lcP^8hiN^^
zPHgfAT?YnkJDJlEh~#2630#_n<9<3d!WN+etJ_y^F9ppbn*u0JQ;+#=)_%mqAN>uL
zf~adPE~=kRCKkuEl$6lhEWZz$q)`A#h{OAPLG>o(~ko(|2n`A$wMbVCc9y46HgY-SgQaxGz`(^J|a05(BH~
zZjd#ucz;9bu
zF+tJ!(zHxb;#^Q~2ocu7MEzd3TF+LkiA+HV{>W}_Af{avjo6%EbYyf0*=coBAKfMQ
z;jnBfwoK|a=MVp-;;Mx`Npl>Kc#@larhm#;zR|5}oi@;4@)5(|N=D121YEmT=E^hN
zh`jq37~d_2q--6)cya7+d3*N4V&ZeO;~2Nh!nTph4KUArT(EL2cC1QggsrN;orf|HqN3IS#z-11X@3ik
zk*QV#+!~Q9K@#ft^{>b*jm;_eE3APyaZ=}bmNptx!Eoh&d^;(Ck~XgIrqsH@Ig_@d
z_F&0GOa3&YHwb^7-9ZMbou|`3g=*Wr-Ka@ooJE;M0lQ#v<&_ALVMtHhN^})D^I2%X4P}DFJtp;r+@umte6NC
zTq)d&ydAxp;YHT(nK9>74J-W3N{+rSaWVaO1EsLkWHNF129*ie8;fc&{$Hqg>l?I?wDe#
z8_uwJluX%)@+T;+zCO)HDL%tg9Jny2h88PwBkPo9N4iL(`4Pk2_Q1JTmP=&fFMlqN8I-GhW_pq0
zj5RqjJO4~{#{I7>TJsQpBe@s`z~MXjKblRN3k@@{dVdZX4JmP+JL4^paZwLoCN7Tr
zw3wYXelg=iNYuZOWu@|q7@!!vSt2u&aniy)6T2ap-y;I{S?-ShH-sCv0g_Cv!480I9`*dmX8lS0tay%hIiyLfp+3CEC_xWKeb?!WL5Kx-WrtPYrN6r
z0d8#|E`MMAhPM`IG<0efyspB`&Ho3^4&27Kc83D)Gi-Fv2GstN<04}0P06VqB1W12
zsYe+MzHnX+Io&cs(M7bBCvRlzOR?R-LL0R-FN91;>9k#57zJ8$nYa8eG4paSc144p
zMwcIJ3t$T6ibYAt#$A1k#P*hN@;i6Lr79;!Xn(AN+t^RA7WFrs9Y`c7i{d2xp9IjI
zAwRC0QUOVMv~&o->_sGfzhWIjO_n;{61p35X$Lt5eWUBY2k=Gx)RxCh~VKLZD9jfpl~Wl|13zof{C<8
zNNI*F_402crc|^L7#&`%VMwQ{PyZix8$RP%f*X7|Y|5V0Mu|a8?$ZW`8oZ
zhYQsXSs?ABGLk->EV;Mak5!h0d8#%nazt4+-t4FAKxU;X$44<>QlmrA<4(!s0*$&5
zjU)2NIIU5{_(U)`M;=nK&u*St`imu0rQU_7MbOK-im65T77+Z)B$KQhV5Lku3@LVi
zsOplAPCv8Dh%Kz}>(frl(wvF&@qc7sv{jx*G#}=<>h8Qw52d%xUrRV=V6*9*KdxeG
z)Z4yfl_<3%o7Q}pTkTvIMHY`lAg9t8cVcKT@{1~})EpkG+6CWo`kqE;GLLBj({gA9
z6gJP_g{B%bTMBD$CSp)~uxi<@0)G_NaCe)o8?l#5Ob-u3$Q5B}&2j!PoBO0d+I^B_S(`wNGW{0fwais)*0t-;EMU-PBx%|K+IA^4brsH
zJTC+Lwp&3dCJ6MW)+g#vtbYcKL7%r=9YzI*k||f-30CrN=5MG`o5$1juI;JwA?p<)
zcL|?di(h6fR1!AO_>Rn6jAV3r7s@w5LKbw1R$Ln8Rw;O3P5XE3R>ml2
zk7hv2Un$tiTM66Y{DElr6z4n$F20<+hsYwnjEW`krEIq+<*Q`Gn-{#O_rPekx|vFE)mmu$NI>!FlgbR
zx%p-t^A{AJz6gS78U3fJ
zfu2yyBXA)>HyIC}#xKRaH4;`4da5Je<2dM;H3LZ&pPg^dz<(>lb%c
zm7TqBXjDIWATkoOUh<9<5M;=SUVp&k2QcC(JlZz;_ED!ixxEEPaA5uI%1D!`(
zFylMvj$L@Bcc$U()O4cnIqd&^dKsKDUnZOfg3~~B6cjBqp5cbOB)>8=d+EHnp}D(CCy^iY+8p%&_Dxcq+YQB+gH
zm6bm_v7*N^ZvLyp*nLmIeG@kBQs@s8E4wv$&Pf5toDyjGD&c!PNf7LaKKDCI&#H1=
zpm*Gz7Jun{q(G;MdfzB`Y&}^`EF%q1Ue=7DQ{6nJ$ajl
zCFk6M1yX&3NOy_JP$AQns
zFgl+sZx|?yH6f2~Za!ja4)q-!Dtt`rF-bnB5P!&CW-ToyAdQxu2_u2@xIxRj@%Tj>
zgiJ@Tku_HWFYh8%R9xd6gy_4j6Alj$gA+MUgxKjir^3DtyTBnEXcOKn()jL&R@iG+je5
z{lOZb$P&G6jrPxqPKZqFXkrQUr{J(F_s?~XUA5Ibcf}m?=J>`oTs9I3lPjPALco`P
zfHWy$WOZdGu&JFs^}CvK8^ZR_{XRdb*nePO5^~y5DoVBqk`V8hbBza{>qP*NKySaw
z!TMMXUK;(BfN33(2N60_P&lbSQ%7e}*J^G}12s_T-B{&@D>w)7F|}@s6J+tHWO}Zn
zgf)S(0hip5a_Q5>JHfKn%QBK>()P-#sV6lgdP!eGRoA#mn!x3G6Qc}np^;(!&OO~m
z$gY3SGUj`d@ipd7h9>sOy5YQQQ~{!^Yk&n_;kcB=&hQb{L_*lP3EgrymbVeQ4A0tI
zF`5fe1RHncYd*awM~1FdC!!;fp5~}87#ize2XeH2mQxl4d|c0+_auPQq146IHI>+A
z$cDto;xDKiC3$L&6?%Y?e8pWQ{nbKnMans557svbEITos-tEhVu17El9(1?Db+D2u3%mge*}b)o-)12n(c
zvoO6T3D~wH3i*EHrC6pPrNY+u46>bfv0{D$pmNGOn(J-i$+5!7$OgtmE@Zx=(UX6K
zsq1z3x1qODY$=`f=A|J=#M`$Xm8rD0&9O3X3iO0?iSlSV@F;8{1P+zVpT^~XvWEFo
zNEZ`Z#-L2KO!w)COJNuCXw3SYrgY+DZjsR{+CSNe6WV>VFQ|~x7!m}}zQsEN;{0dQ
zB@Rz;jnLDxf|fMWr@WPP4d-3^hBJQ_Xl&Fcu-eM`g2ahu73gB4f2RRdD?$v+ReZ}
z!|~M!w|4O0nBubBfBa|p(;Z7obgxOcescB11R}LwM#z%!
zxX`VW3!hH3H;XZ>B73u2*BOF5)-SD6-^VIFtfAu(1sJhmTdMRY@PRh^
zYjt}gd|sYbTkVg7`x<7x>;xL5yZ`1gN?eofkmxQx)j(}%&>2KZZ$u(aH=;>boD)*g
zy)D5+fuIYh-V%u{&m$_7Wwo`PVL=7%)@puV-nejPk3(*CI{0D9Ii7!WSu#|&{p{7U
zet^2;zq?p?Czc}1mGJfFGP*$s8RD5iJ(M;Nqh14{qjKD~Agl`_eI=ov60Sa2X$Zjz
z0+^m`eK76y&2actmLNa$GGTtr+RV~*$WYCs-I8YA7(RUkzSY$JfESKrR5P<2
zP0zs({N(3jN02^^{!D)?Zs>84dARZ)v(XQly?o9hG^I6cW607gx``$~=(Rn?KLLqI
z=ocZ*{~bD|bE^(ALosD{3GfVCq<-D}gY(b+$+sgyx>SD?0k=Yhz=Gz>;ivZT
z?z^gSVaML5A|b_|>pe5kxSE>S`LsN{{ISK7P`$N96!dlih+IJPg}0halBQhwDcibq
z0aLUvBOoI=6ulo5*etwzQj+E(?9~~DwsgBm79H@4_bA!Pydfo#6*@&@%k3{vQ)h*f
z1uVmJn4LD5qAP!g@DCQXOMiEKMm{#zedaG>$Iq?M
z4(;48w_)Tlf|*(MCpGV5>>zdDb_gb3!799+}DZc
z+1G)e5cyYkFf2o_O-U0>TK8M%(NkGd6Q?i*y~aa1OUd?-J%R9ME$Ei<6pcH`yY!uU
z^m#7cNXRTgEOjEQ$3etEQJJqne)@|&NStkzgNn1Bh`&+oZ#
ze;XpI(I0=f==d5?n8+9{9BK&bGKoPR$X=kS{MyfJBpy%>oA=X7cse;F2GtHAFfLCs
zTK1|b37Y)v!}PP@uQu&Ux4GYvxfApt$~2_IX*Wk(ceEJ{c0QWTtJ3CB13b9fa(AO#
zAodhK(AT+wu06C>x2IRbzhXjZ^1C0DFp1ryOK5*3?rW}api@AfU$w*X_q8enNR=Re
z(g#>nozzh7M01}d`%&4rzutNHCh3)#Y5nF^^GGH+Jgd`vh%bIf@Zlx^CTBzn?fKu#
zjz$*4R;iLt9xl*XSl;T0cPhiXgzCyCxl_2(Yhh+SD+ZRP+i?Z8Tz5Lg7<~5aX3!`|
z(*1v>O+QJR6=-1Uxs2Y@e8Wxj8@kB+j&5Vn$vR}UdigjcR(h^1#9xtt!C`e$L%#_X;S!;ISRB`w`bTDOIu70mbgWP}5
zP$~TTlt)IqAUOS7L$NY$X(J-drVBYiNgIQ~6UBM1fiWHJH_A|W
zarS!q`0DgdJ_BxO*h57cf%>*iTNrCXahvp~K_w_C3P?E+XHeK$kKDNiKX^bMkKW=|F4k0xC
zqkStQ91rqEPN{+rH!ZA5n){q7S?6oz$R;=bdc5hd7pDRQA+&i>s%{20j#`v!FA2)0
zo9Cnhs6tc%zv#x|6mwP|5+m9KnZ??Y8Z{7T*l-|Td~7wj<%Wbh-%t!qHMPET_`39V
z89RtnNvnx*4{Lhc**^t}*V%tw1IwEU4`P2jErr+BY!tEvj%q4zL|~Nn`!0NO-WFPX
z?rD$*|E$rf?k+yvhSB^bp<1H$)1&6kL(cr2+j#?H6%gLTTD_V(cFRuBf%>clI*V3+
zz1HP^w1tPyR_eQ#jpX%qjcq6krED;^UWAm+IKiXZLtHDcRtWJ>rw4zaHt{`dPfT7c
z&A!C-g9r!UE)A|`{dLDQNiGIuteYNThU#GI-GT5?qZ2h8Z=VKS4BUE1+GKlCFdPNJcLIUV)xsLQ=wo`4P&VtNAW&oGm~-a7W{bH~
z7Zw$Lo3wQLH~K%>$ZUUu0`5PcpdS7BqwwNbYays#_-xgZ^rgH=(=RY~wkCMDI+=$`
z9d`Y?TWKs=uHUdsIvYD}{NfIp7K$2m$l>(GH87B3gEt(Xq3H_zS2IC@#;-ix<%%;8
zzyVy(^x%Tc1*(H+BG(EEDtqf*G~O=8f*A^To4dsbDJ*GIu+)D-dG2*3fe@RuLqXc*LDczB}P)ef3?tnq;?MsB
z(lKw43T19&b98cLVQmU!Ze(v_Y6>?omm%>26tjr%2^Rq~mm%>26}P7c0VfFoHJ2gr
z0uwAUH!(K~FHB`_XLM*XATu&EI5G+^Ol59obZ9dmFbXeBWo~D5Xdp8)I5stx5&Qxt
zf3$gJRGiDUEfPGqLx9HJ-GaNjJB_=$ySuvu3l`ix5Fog_1a}P(;C1%C=VYIA@835V
z(9ET?X05NP8_0TWLU01JH5Jw0KUINtApYib0C`oA|92ly;O}%f
zF;yW|Ek!YA#@}lIFaz9xPR^FU+x{0f3J^2EA8MdfGbel7zZw9jEL>b1co-So-Q5|?
zU7cMR?48UR9BlsLr)pv83~;x1f3gOEUY&q8z`ulXwKE0D>0$x=yTI>V0mxdK0PUQC
zzmvr6|90Afqy!~_>@NQ%1`@*McTJnW!vW4f;D5BSFmnDYS5{F`7GP^+Y3Bm8GqN)Q
z1-clyxH|)cWs?(x&f36OUYIc^6u0Sc#{|N_~;Q!dnfi3_h0MHQt^f0kt{N3=c
zdHHQ-{%r;c;p^>S?*K3}vT+9bTABetFYw;ZMs7fWi<2wR*ZZH2|3vW2oB&fx6Bp2^
zgB}<7zp6{wnb`xl|2Bh|{LA$}LqPS{qe=~WPEGCYY&-#`Kr?toe|dWs&`41I|9R5>
ztCzT|jg7pKEs*MeY5L!RMz)qVp8py8Um+U6-)yPm?VW6mZ2rq<=`3#P0W?*#bTP5`
zTiU>v?SRhC0Jgu=
zfFL3Mi!n(3-+Td#e`4Aaa?*;l|9foyiW9Rlu{X7}GY7D+a{!E-oQyo-nLx9^!p;ux
zW(EzmDbVAuVF55Q*x9>)QUDIFF1`RWdnfqc8_K~BU=;do`WtZo7=`~JP5`6GAH)S<
z6#av^LB#%zIGF&9;(rhefKl=fVg)ct{Xw99GJg=LpX?t5f9faq2Z8#@|3RRB3jao2
zpni&f5U8K>9|Y>B@&~a27*+ouP!kAk2S^M61P?LP$AK*CvgI#>Yh
z{t*GPS^fip)U^5s1W9K74+zrbpX@<&ZT~=K5V=2<*+Hpxpr_;?e$e{u|BwMC+y93f
zv>1m!Zcs4?BhW=+12l8_V`csy>)(gczs#(lUJgK~f4{D_e}{7XWwW&Zvqx+oaU5(w
z*Vdn&AdBN4_&dkZ)!qg8pXHgEK{EbBWM+_Ne~PeynmgP4bEBD=L3RG5{~iSw3n$>8
zwX%a2=i+Yvj|5N+*FVDus{hxe?rdW3^iPFA^1J;5f)sTB$MArPdi(=|x_bTtg4Ff;
z!x>b`e+%gJH-rB?vJ`(`+kf3hOn=U>|9L<27glj`vbP3mSek;~%l#un&d9~d(nE&{
zbeS-N?4YlI{igRH4aoo8ga61D7Pj~Bre_1qH$4kC=;&huDFT`ZPT&8?HTnCk#a|a0
z=&=14{(Y|lfIttR3H>BM)-SyP?(IHCnR
zBtU_Zz}A3liV}NSNgh4F)Ihs3P4W;-o7auDl+Tm5rb>dlMt*XB_()u$jQWL^X=V+J!m!4qp&j~q(D%S&N(FvJQ|)E*{6rU{dUV`a
zQ9FXsiFEEl_y9)WPxhhGTmnFutP79jo;4tcoeCK-i3FZl;@s^TSR^$iLYWtQjN61d
zz4XM*dgG%dsJa@IE_S-wf0%Af#*=9)f8nK^9cJDk(Bi?5;Le*$BL+ywE1N*aS*$_4=h0rehq=x}wtT^oDoKYHHj3*5-U=&>Vi^3NE
z1EV>LJ|wCf4uKVS0;D#_RrQ7VVMsX+6F@jpMi@GGK0+tGPGVM}$4PAC0k-J}f7k(0
zb=jR3SKsIs2ly3ATbo@QA<_rU`II^uccPF0ar-qFE)j7Ao;MtG4qsD6?43`e*2oEO4c%?HEmXud%zS6sjoW-(@>|{l_#$7gX$YNUY
z6Z1|a7)H`@s&0+^I9;01)ep=2?TXZEe$KBHbRYJoP-o$UxJfxMTp+Zm0tM=uym?&`8MY(E-j>#4e}zy1{*vqK
zXR0tkNrDaaGRNluB|Og`7Xp|cDGZhbC1lK6nZ9YmiRl*BcGr`=u>wB_f61n$~00d+lCctl@^kv6p#GT(*8x4G;Nw>LQW3!A^vQg-BsON2v)@1>qR)i;L$C
zex6;*4+(gLn|F{eO$6OBf4)%s_y!Mg<=?<>zS|MTX$a^a_6M=xOSSOle#{vh0Mk2>
zEiMX*rjMSemb$HZas-Fc2TWu%0>tFt>3n3kSvv!(kuN_GD9sozv}!N5Ik>)`TDVq3
zTCE}<5P!5Xq$)(Qr7!zf3e!JbVMJ{qfUX!(X%Y1qxU}_V^T`7eQM|(Oky8?144@
zy3cl0+tVYqIc8_zt!ivEd#lI+bg0x;2wekg9K~9xP49(|XA`IcWuw32ZV=ejZb`V6
z?Qj_%5z`TY3Ff1Q)Ue0e>#DM>RxI!8zK=y&&KBkU&_=DWegmRtM>qKd{3pQQ)($Sv@TomLeWpR&mWL>4k?f1~%sHx0&lJ5HOaOvMJn7|JXI
zp7ScH0>n)G#S@MdiVhX8BA>s_Sjj>=!g;`lL?}LI^FoAveOjd|cKiTM$?N$%D-xZ9
zvwwT2)PD`hfAeZvf?gQT2h9Spaf`SW)Fj$i>qvQQs}p}$;H;G4r$T-lOC4J4osV%$
z=c5E?7TDrQmqbYv82$1q>L*-I=laCkw*Gd0hbKyumSv>6YA?J2Q|V9eRk3tdP8?1i
z-uSd=48h^V`X@)Xt?x;ch~YFkz@1V-5!&<#f6E8+e=@I|_AR>Pl7)OPSS*I6XB#bf
zd>fSR>owINpDd1GXb%KtPyPYTC~PEcKm&r0pSR1r7b%O)(he($+YL$ZB(kIBxpJ4A
z2U`n&XwoQ2E0wH`sj`mjmy^$)le%+wdCI#2)8Y=p_ArPs4+nPXwF9a
zj%76M*S<3^C;p8g8@tZV!NF@bG28A%cb|1>9h|RIg1fWov>+$?P%cmRPV^JuR-jXT
z|B)$0g{Qm3+|G{!cFri2XTDcUU7sxX*!p!^e^rCqw&YlbBZ%w;AjMb_0wgJge!aj*
zT7@%TSAKZW^H(xsqP`pfVIdnv(z`WP>BKHI
zegf5}IcD4NdYbPQ4FUL)}nV|&?nz+s^z14z`L(KIk3k)3r4#3NOeYY!Ir)3rT$
zm4)1+5+%P`Nq$Ygkxd$#JA`ER?Pr`FT_PfS8DGbSx3PqGEd)O2zE9dPb~$Kc!{<>g
zH&Vm2#U!efSkRtlhj5#C=P{1$fAy*{t@DA*&wrSA;w@a-)lp%V>}APJNPhluo$gKC
zwlXm|7cy>ftlKiV!$ob3Q&ybmS%`DqzzsG6H-M)1bVEELv~!>=MJ{XKdVQGn0=@@`
zKdTMrZ2j$90p->u_0S?G!?n~Cgn6GJdLB#h2qRt)zGk*&uY>^kJe=;KaTi{MT9{TsdbYkG!yUyZy-!HdtNxaBo?!GeP`=&j
zZe#hH3Gi#PVz41eKM7O+kr%Yoy0koS*gqYz543r;rNOiK8sp_ze+1E?{SMP9ZIS8m
zm610Y>z9zb-tBe7S9~S!i6E+;yjl%dB8}C@MR^}4Y*QSQG``|0J@hdsFcj;&`Dr@@
zv6nG(*O4%l%D~M?g4=|)&&@T8j`XmCY(VkYLeo}yK7k(F58i1l6fAk~I98C!oT9G`
zJVp+yzMa$=gtH4Pe|&{m+ecLDETN<1We{0{Kt(r+_)fEm@0`#IHnbsDtd#G_Q6A*T
znmNMErBFzbg>C@kp$MVNU~#9zJPA59XM4+`Nc6s%e!5S3oGghcCBkkz<$!R-FJF(`
zA&7gy)*fDogA};gL|*e(FxNjrYG?t*-iMEVa$P>=b267!f5+v1VVpMk<;wzRW926g
zOJKC_C9upXkmWD4@@rnFtNWBx4cegg_0T5z6T{rmcQY|e`Mxg*%#r5q_A{k$!DCxi
z@WiW`^j(2?aAMxQh&~h++h(kojZD)ci1ge)DZiXPTOpi?6c>+p3EB`X|}RlbxP!p304>k$7=MPe}wP;u`oO_5_2A}*b%jE@kg&b
zE^I#psCtF`5N3L`2Au@j8T$%rD_H6!wK2mve(Cerm7Gb-aq7?uuTHlhdg((p{>nH}
zxpBTAD`!|c`>vRqWIew-c)#C&en%^=u&<(gXXh~=eYKUu*fqP@X#qiTH5*1q;F(4h
zXccZ_f8SpDlM)FMBXs-YAimHC6&~H;ohJ1!9Fpohh(>6nqv8Fz_}LM|T0P6-ypKcp
zf*!(fNNkxYaU$Czy-l1p1{|=Fqz)S}JlOq~QV`Q`jRUmFUu*{y-iD&qZJ*hoZcO1C
z-T5+)$%^oVinekEf(@RmO+Q`aN<&Yr)azHfe;j(Z?8PXiznuEHrJ$1?90~x1zs}`B
zBjYCzj1H{H>Oe6H2;*wfi=|}8@OZ*}BLC77Xri~Ec>As$W8i3ZKG`ipgZF7zruTjC
zo!n0dS*m+Hnv21g0cUBhy$nn20oVGUWzV%+{Dlc3!H-Ft1-?;L})B-L)G2Nrhnu_sy2`u6>&m6pd17gp|9)*a`*_)Hd*>~J~Q3I!hp
zW#20U?|TQHf)z+vveM$XOjftJ=CPh8X&s?rCeqzT4kfsL@LP9Q#H6#Ul%tNU!-8c~
zVM;CVDZnL7lHOy_Ax?D;Re8}NcKr1Ff4YE<;?_=n6cd}E#E}fAFq6vfLsL$%}
z54phJqaDUj91azp#r07WpKpeapE-?oT!U&sC(JktXOyE#@Oo3;YPPZj*Nv=cf8TMj
z!{V9{)dVsgkyTJ!SE{Rub*Gc-JUsm3>(kGYX~dr~&T(CrLh5md)ABCri$fn;)w%fb
zsxAT94k&W-&W~Z&WdWL;^Q~Tf0?D-2YFm}v^xqA>^{6NjP`qXg+hGspSPiwFjPyIA
zHM@QB2ShxjvyY1Q$|PocHf#s@LZvaTK4D
z(rvnDqVv9w?94)M2UIC6!-+J-xfPaW1?#obC1KI5ZbOW>Kn{y2$SqP^#(ayzl6}PM
zO}jW@sAB_bKo0-ursZ^5LE%NUC(}V0?nt*gVL3dt0ozKCIYsZ8@_8MyIE86%_@
zen_mB3%=S+)iAS`>jsC3@f0MT(`%GWFggg(#P)9brR2l-LTLG6mE%4@-ecwX*Y$%<
za~f0nx|!g&vKz-cNtYG9e_JZNtM_$k7_F+gVtMW7{pweUW}P0ZnrMmV9A5=9LN1l3
zSB2MY)MwK5k*;nGiuAKzEKG>gm~Vt)JM#Ef&-5rP^9Hj+Vi~inJ@%`Olr{HFMrYKX
zPYb;FpOPM_-z>|=rqy+CTCLuxwV@H^&Jvp
z?l-oRPYdYb>O8xoD^X@(b^6boC3TbmEe;xoXSb5w7bl!7%;~%MH(bQVT`K9(~?~g!Wt4M
zOEUKyDbiM@O_Xruf6qmGqh!4+apFmyd`@Y`X9uQHnTtB_?bLyrxGb#C;;gStcKxNl
zbmY#zy`W-YPzJ{*e@E66K@E)+i7V3){ZxaO
zQ};nHf0D;4Po#hu66+xd^>Za&MXGF!P+<{~^PVtHS-w!I0bfwsG{fg5fD@6_yk-j5
zJ9&fhO#$=bE_*x1-6=ja#&A}RetT@nQt~SY;k=2$Fc=KebC~ngzO_j?9#lRCuh`3~
z9DGqoUT2P9e_@DBR^3E6$EM&^sB&rDRNIf=q4clr=ebP$&w&Bb%eVgXQ~5uXu7p!{
zqIGc%e>|hU%wWqCC0TApg%loqL9jE0L6~<5i3#m&xmi@$YOgAWXFi(Y{bflxJwi;#
z!vuDKcR7-7Vo87&VASXEWoZ=;(=M0H7=TuR9S+~kUKZn$Sm
zr9_(f&^*KL-v15aO$&!uIwiul?-3Fy4Sb<+vC{q|!^L6j+!|}_;jq4=$;l{;)%tF;
zy9fNme*+0F7zMrs^GVjP!!cUe2GB$mHO1y>5fJyDxioUh2@bXH>j~AZ11?`7KMZ?@
zuk;S)M>pI=-F-|9&QR`QX~PQo+SnBZPdJozJKA|laMpMF*mw+)69xQHexX%mni(b5N-mmiO{9&+hZmzjtvYWV
zPu7+@f7Q%|P2Y*`yIrOBT2!gf5HoLEg9B`|A>R5iGsdaT
zP4jUZ3or3}K&*+?nn#961rhPEI|Bu_e~Hs};RnrmpVaPhTN^`0%DHK0hZ!9H)Radz
zh_dX^pH`>FDl6RDY8%U_-rs(4j@BSJI4AUv_4YfqeR+4K4)yU9juSca!m#PjaW^nN
zZX{VV+d%cF;yb2QRMi`O5!T(&!j5lf^^Z0jL}5G)4dedSzt)6zzxncVT*^pYe=Rx0
z=@7ZQ&t=O-CGy$mvmSJt>sL_U-$(EJK
zSa`Huu7p|_yzNzbjjZUA%PVGWrhiCKHE0=IV*PxOdFf^tJpz
zCKC&xHon8(;lShEx^Ov=N49-`z}m1tozP^73p>D${$4)?i#a?R2YYu1f41kz{vIp&
z=?b#4S660N-?)Y$5arbx!L`sOWw49!1_7HgRz=04=NpvB=H@yCo1ZmLjcJ9{j}f15
z+%-hloy85!#L79HwX()g0Z}9ka4<+Ih?iLewpLy7%~Zq!?~bY>$88gp{7SyO>{5G-
zEk|ZC!>mXe{i-vS701#5eW&Va>SQbZd>hL8yBJhlP$8q_l
z0y>tZdR4;y-H*59u~9!=?;Ifo7mX7$84}|RC#fOBPx0Zio`~AwQ#9i}$KaAwA3rk~
z6D=qf5QuO=HPvM&So;*q7NF<|UGUotSq6ak{vd}(hvf0^%yO@~?A-b^~1
zMJVdoTs7j(*ukePW40Altgjo7H+)m=aFt5ma~n8tjAt<^Obw|mMYuBQRL*C~&g8Wi
zj$ES3;m);#8>x87FsRwrw0qwvEW-kkRnI0b{fX8jc)H^;R8XwCI2sa}=yM)qjKI0r
z$gP7nw)9FJO6|dde^JcE3h}J!*cY+~zs4$Ee%el4`8=JQ3C9~KDifVM|A}z{ys@8{
z_=V*%TQVd)*9bjm*?JpI-fvMQjQz+53gLKOFwKv9a$4NpE~I}e>ZdO|SiofTC_N&@
zr;Qc+t$g2s{CxU-D>EF`8-Qtk&>R*#3?&!}1{TCXnFgice=1_Y4iWm32);L}pNH2=
zwtaFeb}*{C1ZRtM)klgLY$9Z$H>GSJ|#H
zzzEM!!7o!Ve*w9E1cGw8ne|0i#fkzyXSmeR`OONRt0?W(6z$SuCI;9^Q0^GjBJcKQ
zD&-P@5)oY`aSL-8OpJ#;)F@>LF6n^c^^Vdf+JPXje+N$KWw@UU=;ez~RaI`;kKjyG
zMINbQU)=aHDFk6A!-3Lh#9|!s>9AmgQoT(SaJ*@wv(#?XMh}-G_*QE~*IM-LJAzrn
zL>H8~iQm+4rlntUXIhki$B3aq9S{6X%61l#`n|p^x?vqHF6N}}Ht8m93Cp4rP@Z4PO~
zo`m~dx|xXZhaJu>!Bo2vyzTqVe4e8$lYsF}xAx$JQ}aN!qa0J|oX3nBU+4`mgzm4A
zuC63;m#c)D-*?)+K}Y+<(F0NIv|zoEy^cb?fA&)^W6(1s#F!Svf%66Ja?Z#PIcN(I
zbFH4Vn@J4f3z!f?I@5!CbQ;udt}x5IF^F1rC>fhV^gY8_NK<%&?-a0GQ4aU4k1Vni
z0o#VJ%>euFn`gpJY6&+idVN10M_6D3e-Q$SgxuJkj;YLlS6`?*NFDVgFBI}XPU39w}tQc?v}Kk$akF-^ns&|dDh6a1KMd!9?wX(TE-9t4~6)uPrRxYhv_F{IezlYlV${H{pynEe}f7IJal^tIs
zM^a9gP{wX1!DzPP@L6C$@dI~7ds6h-=((+yvd6*73x1a^$MK=;J6{tc5)|ag#`cfX
zm8ru%tZ}1x_PfvMRrYQ7)~Am;#s=Akgc*`gID<$exR&IuA`sh$GGAj()n4Ipfut$6
zy~JLF8R2~k)P-wJ772Ixf43MPXLiZcglaMQeI*T2$!-F6mIE*8579US2ME=C)k_s-
zRqe~ixCk{C>yRf6$ZWJ<(ib0APJY7Cja#Ozo%8j~uhNU>jjhX}_GhQj0H0wA)jU13
zt(U`!59uUXAt*l~t>>Q?#asS|`3*E+l>1)q)jF+Uu}Qf;`Rl`G@8q|aAnZ*n-W
z(D!Rb<_T*I*mC>-@khk*T7s{)%JE7PIgrPT5{`inWHZ8hxy&A2O3dmSfP135>L4E)
zSQS@h{z)iwdvWb=XX~|
zdcFs<+IIx8*jCbFDUngFX2@`>%kvaq|WlIJkUEHNz~Z|Dy4xfi2yZC?<#cAJh+8^?;Yr~VMB
zyXEW6b6Za+q<1puG##&VWA>Jeu47|EfHHD!_Z>>&6IpGJSxwMbm~@-Ot|logbN(fM
zo<85X+a}Aqn!Yq5pgbCv9$`NL?iq?
zpHUyxMHG_3P_A>{XJ0j56u=4A8OtkQcKsG~{usRP6HRA_Yn8j&HuQ4J^Ux@QTi
z`E?};f3a$2LsMS6RAWhkn*FD~>NM>I7;rIb`&l@IC^0nP=2&DVu9yZENjwDcmB
zX9JOT-zP6?Q3C2M38uVx|K(dmwBCjFg~Ke~GVy2=k_*C(>wP_-o7WGA&VHEIqje?#@9e-k!f`xUr`5Yr|nn(Z>Ih=xAySmhaPj=MTUQ
zA8hF|IhhZxgsM~tAKywCOuluolE6y7-lTx-hrJx&KCCwW`OA-RdEOoEbi1
zf2}1nKZUC|f3km-8@Z_G1AZE2fu<hLmih%T9^%p5eGDRU_jOdjo+Q7}vDg}4n
zJ~mI$0s9$eN$rG(W}r^e4FwyDm=PW?yK^bMiSQby9A}wXwvf*6UzHf}@;@+5^L8v{
zZlm(I9&ICZ(^Ot7OU#nP@I8%kzdR^1fBSAfpNjj!rgO&*3JQI1Pg+QNX%k-}jJPi7
znI9n>supe(m!7n4Ik6pLikCpaLIYXpT
zs5^$+R%FhKDp1?q=WU7h?UOtR{J_&+uh+1~3eGsOIZ}NMePLx}*!{w0w@vBAz>N3xl8m`%ES(A%JrciA;JhO%gN=Dxz(dcC<
z#n443)j2zAya^Tz_u3rp1-n~`e?LnZOG_8Ml-7=RaPrtE3ltq&>>X&tpN2|D_VOBi
zwrq7bEyuewZ*C`Z027rt5<8@5u-(^Za(y;ZiWe1^5WLXK4g&i^?`(0J8tHhy9;=sr
zbaDOmLxX!14XPr}?Ro^E#W7vq6@3cRDtkK{*^`7_65g%m+<`m%#%BlTe;0UydYjzn
z?oC(I#~L#q;m>84l-I6U81?(Gk<0$obk2lAud6Y7MJ*vCS9wCEPa#MoZ1NOA4^-nB
zFE3Vfs_5)eJFkca@A*35Uvlm(s7Vhx>y-`wyL{#1Lv!20Wo!f!dax96RigOk?~=8&
zt&)NS?oRW`X$0fw7Eht_-
z46E#(+v0lFy!WfwnvW~3#s%E&rs8`iX7ZE~Mb|d-yf|~0rfp2fwfs=pbl|RpiD_Uf
zgypDWGvmP^D9aX~!+Uquy1g{K!MRT=Y#lX8LWgU0hPYX3as0r|f10T$$0dN#xZ|u)MR-PRqXP-tu5l9PQa{+kY
zjnr{SHFs{_vy(kqwLnHdpM$UAQ*10^8PGiEH64|=jQ^twr6^_wXIUd8YOZkFZnbqx
z+`N8t#8W>kJKIsie{#g7@!*^%ifJ{|0qqrQCf-Tmm(SZHOu`!T{FP33&%~u*PL14k
z#L>eN%z!)$mlgQCas8R$XxL^Jq6or_#$MC{FE1(rFn=Pu-7)N~mwNLlHRuZ4Tsnwh
zeP;(Vl4P_3oPC(u#gF~*&$kOu@OOUllj@_bsr_~1;Ge8le;?mlZl?4*^*=JgXYcm2|-168wxfRC$cB-DR~nrxq(Oi=7H`1jYvCQ`H&3f0Vs}Y$ZyiuxiV7&snz7UiphC
z>h?^h`mdGNrM$&7umnudlT3|<2PF!A2j6n&S`qWE%03HxSK;1vBx8nH|0sgekVQcU
zdykF2$KAXw%ybhY!u!>+KSz`E?fJ9tEm&~>9@@FS9n3z@=;tC;pRVT4euR#ONJ)De
zKhn5>f8|4bc%hT2aa+fty(Zj3GlXv3~cjuCGBorU*M
zu$%jY&%wuC@bW$YFAN0weIc|1*MfJdFm_|pAe_>A4Et;`C7SuAOCG(?ih{*6)?4{rN
zp&6N*0ZJ{veCF6&lj9-~aMvf74pXfvc
zuMd(7&SEimbDCZv6)!wGI7HeAAxTtqe_vTbYtNQ^Zp~4SVaV=Pi=uXL)OjK+=}|W-
zIWeYI=H&*ILpXMPH#B0+^Dk?!62W)Hs~-&EYd^;twcnFa5hhD;(I1A^4LCfb
zNQo0FaBb&Od=j4T|EOJCY`6SS{$sg&vRfo%O9s*Yw2vsSG9y%iJe*3lM*L+{cOmw!
zs1i;_@3z|$NOX|j3p~Xjv)VFJe{bDKq&6ZeN$^NCp;?+?WJzbFJH2OEDioi!Tc
zXzVTyQA#KFwWBrN5BOkhf^r0f#se;ZiG)hJ|}McYQtCWhDt
zm77q(aoT7cqFyONf(jEovl2!&X2yk0;)l;ttrskD
zpMLD$Yc9pA-8R%7af5|=X0a=R50tOrusx>dRtqBPL#TBqW;>M~Xnvma(X
zpe7xTv)K8@)AJW1kv>_r0h5@Rbsk$~T`61cN~hpCf#tKv*2FB+0w>-|PtmG(PGG)7
zL+ODm<#R`;Umg!uO!Nr8VELL|g#>D_vvW|J5yCpt4WpBlHi@*|q$s0_QUGomFFcDNuOjIHKS030%d=-D
zp&?;_Z7|-7Q9Y4<|EHP0@7r>C=AEtm$)-G52|(>^;c!(YuGV*X@6Ow($nBA9(?vVN
zMUz>wLZ~z(GfG5I410~MEOIl=eWbP1!w_?^Fn=|DElR4roZg9nKg{mVq&TaX#u#_L
zq~hZ24#f9`ASyzvb7Hk%jLIP=yB&I*Ie$#J__$BH@cL5)`I*yaqchMtFW${ba3+)?
zW>W-c6ivAW#|q!!MizAwYVO~#Z9rg1bT>RJRh_4MU5*jEo}+Be(UyRNY}Zort$q2V
zpMOTp$@}(wm8hj{z@!*^kzVR;|2j51x$`KvyQG4j3oTg7C?k$LBRGn}s~aUhe9B1w
zh6XwNybh83cm1(ktSSYObjC*w@l=zk(Ig7NA5Vwe37+tqEps*W+f?!I%=kX_YcM1&
z%P?{+=3jwxoQRfhgs)KMKg`PrKB}f4jDMTLO*14+cXY@s2{A2Ut)i)okxOn9LdH`C
zo!>TRbYNG>$m!1o@-Ug&V)D55I<8BMty1z72V^1(Gp(<~pK#a<1h^%+-_3;wQ#4oC
zlS+6(*LR4UV4(771RS8b4I@?gk@u0MNqA7(LXRj(o7HNqECeh_iFq@)uFpW$(|<~9
zKe^NHcm&lg92nFRA(dob>Ckc|b5laLqy;60u@2jckF{}hG*ZtLCvhK0{tRoGgOFJB
zeko7RDyMjN)5jQ@epLc(ZK?S!#TEj8^j09UOv>DZ`0WYF;J&d(1(^tlvUHU-0kI_|
z((d>5DlVb!TZdtw=SUA#rUXmpkbiGd^hj{)fuS`tAzjjYN+*+m=UM8Rto7@J2)6KV
zOgNlMr}ZU@&MRci!S(}V%xiN_
zzH~d&e)(v|n2y#S8IW`$@2^(5yN7IrSFc>BBe}x=k5jpFdb+(njgXig5CLX+dDEm~
zv!OCqGIM|Y$r0amxc2ld#3=E0xXC$Co2SA*G3zcW105C-5q*>1f`4$+AEs6fnJ;3Q
zG6_)R(?wbgvPGm@K-u-2I#0)b!WWKAYnL`!hxTVPqFW2%AP+4H2U26Ud>1B+cBBHf$Og<+-hk
z=ok~~Y%Dd4YoXj3(tjYoHp5e_;FCe0`@39bJGqg+e8$Mt7l$LbL$5N8S&x)*5XKIK=^Lf9N>&cBt1~LhJyGX*d
z9O#DXX%}qn4#!t3u|7t`JQUgN&Nfj>aKDobE*ZDU4W%sMpMM?vdca;>J!UPqQrKF`Z~F;w(rMh8~rumY4b`t(78qRCV8H+Ue0a
z+z_`*f0m6ZYJW0)$F38^6cK1d&$W;H`Ytr>Vt5RSgf+Wco+qm?$2`88F}>K4a5|g3
zVh%tY!MNI42}Y5w^`IJxzEbPk7}D-#SW}WF@2=#&pI(PM7pI0&-JdBEx*GMd#-D{B
zBT!cOP6YQd`oZVwmF8KlvnR%Ld48f$b+2ET;xfKk7JuO`KyF`g3U;K144)B6@HuEV
zsIy(AJFtp}OgO;GQ_@@=%=6T#w;k_W+V1V<6lljMW+^-~VZO;QwCg9(=jhKuV+;8U
z@~rF73u=)<{U}K+IW*&WVM}ULJ4m(?!%FJJIW=5rkY^Zi^yFhABw88x(Y{%gav+6N
zv{j^&*#f~a&h
z(7cQp3^<}i=ddY8)Mnxom$mtRs`gf|#+w40ynmRnV_BhP>X#DrrjpC~QVB@&gVT6#
z)5TUAoUr|;;9yikdbws1LrkB+gde#of5z0uyt({}tjR)PIo+JW?2T%^74Unw
z-VrsbRlf^o<7uvSk>=tg)YE5-DczZRzSw-*8iV;&nG(y+BVYSqEzI^DxlD|OAimHts<2ih81YAIjCjv5-&?
z#`tci%xn9@+v{3RjBD!UUfF#EEh(+jm!(dB3ZrT}ZkYmR$ki;`lG6a`W?8`*`F}RF
zxNU(bQWcvzLyu1;9Ve0ru&m%j0lNy=rdW`~SK*q3nJ{W=F9%PgvM2-9{=K`fO5<@-
zk@yNc{(|X?C7%ZK)bx%0iNl)I%2%PFsXLr-V;nP|hiFNZ`DEx58jpes-
zIT-LG^3p0dSMY0AVwvW7=Yqw^hJSmoiScTo%+Hssc(R@R>!~AG(+$ts>OYQncH
    
zNj2TWS=L}npH0Lq=UAd4wpY(|fAaNVFkYr@;+uXdI0au8nQROY*kv5_oG(k6ng*xE
z!k}o-jW#;VVF;hO%U6`~#5%X-c8Z;xW=m2?N}PPXyo98Ev9QSC8h-mUCx065CV|P8
zY!?~z>8Ba)YmJEK;GIqgJN_*$lODm>T)gkHmLrbI7sN1Vd%c@#x~BZx@A=sLA&@X-
zgdK`?z9@{eXAP}O98_{Iu#xv5a^|JR)gnFE+Dn*Yzca1QKE09obxQ#8_1t}=lX{W0
zJcZ60mhCx=-h~0GfJA~A<9{3Cm6k30ib}<`(?~~#%LpsR=|Br7D}UBv7*MtS6S_hw
z!>)HemNal7IqkwlnGShowiT8kbd5GwaB)X1XV|Z_hBUmBmrUecC{Jx>zTl@9(~8GA
z2mc%?_7FUR)SL`qpvRpeQ!1+)Lc{qn7uJHW`u4XB-J@$ma+Kl+>VMtfHk0Mf(ga>h
zlR|#}Lz-aPE8ZmLs$!?4C;0_|YKP0VjTV`TTI-L7O{#%!GgmBHr;~62@$jmGrgK1$O6de)lm!6)JNii>?Ab2HG6I=I}RSI
zLWr=6u_?qP9|d3ByQkG9nB|4f1(@<)RFp?cy~JqdtVChJTA^3KvtpP+5Cy$dl!`
z7`;2?i@+Xp2wroC=6P2b(F3Jw9`SMDMC5H-+3PdH4YfvK0-AfFoQ*qlhbtX_hyB7k
zzMN~`Y#Y_diQebW#3h}#jK1i0v9^f7ugl-C+=M$5SG*LyzfsXtdvE`G8_}EDf&UPU
zTBzxh!`}3uA%FUXEB!;`*Jb7{ISvZ_wtUW##Iq2@i$W3q`@-Pve#x`oTu+p<8AJ#;
z&9yv@1F0lQHuriuANau!clji#g7a)tm&M_iUnx7V7DvKZd`=iWJy}*iT(&ao27xs5;7Y6`ol~)P%Ju0icCzS*Cu+3seiMN@Ar-kpKI0mBrVxFdm`1+|dNotlK$9$x
zH#|H<1Ahr-P;t~8&w|@9gfH{S+ZJN7qV+|`0)L;@J&_DOwG#z
z9XZoE9b+>5^}XK}XJW(d9_EV$Pc=@N2R-nIYrG9hxdT{-3Tf^(qrM6&O%WZWUu{9-
zavT~eWU~bQ9OE|{vBK*9SWRi57>COp%uTgts6b+}N7hL{+|ejDvIjv^=h|Y-Cw#e2
zN`DcKgn`CR^@BUPSGuXgeMdu=#u%jzsJyHlZK+_bREQ)+Ke&Fy@Xa*PG;^^CH!FuC
z;Mi4Wi&b%SSHrbsB82Jm(nU&FfXg%EwrH_q_Nb+QnSJ8i-LyBS{C{lSLw6>Cf&|dm
zw%xI9+v?c1ZG5q9+qP}n?%4Lc&CF)rpMR)Rn_Km8MQZQEH;tNidn<@z)pkT(K
zcZxZ%1#lWItQR3%W>ceTBWsO?Qr||yj4~v$GI|la%$&|U_f@*}C&jT%_vmQH@_(qV
z8Lzt491~bUZ;Lp_TdvcfN@F}wM!XL6YP}@*`xg=Wi+>1D$3q{Z9B#m0j!S-yFNN8-i^|DL
zlpOxZI?J(kC9^iNmUTS!k|@n$-#Dk0kClg$VC>*2jY`3)Dg#wcfzy}^&VMlZhK^D^
z^A(0g2bZpU)&UnNF)>2SxMg0G8C&$W1aon`3Ni6eH~up1)*lBTZ9jfgm;!ff+oz-Q
z*w;A*BgSCy3+M^JhRBgBEZ9xcyR&l@A(zCJ;fQ
zOoyRvmwBjua(w8oH|oYuiGOJ4fq3!IB70eXsKz)Z<@V;Nthrc4w&4?tgeTK6mevlA
z+b`v_?;|=m*~3A*kKGFr%^iS67S!I%Gcz}RGXWhUu0wM^!dQ9xN?aUnm=ofk>ezKT
zO5b!-ZaaGdO40`kM&89bC9qqy#~n6)4Y`1u+Olp-oOLWe2ORR=gnxR6dvRKVif{p7
z^KZmK;$4)Z)KqWS-PRPHNUwP5a5Vb|l&LUx$HQC$G$XCA7;Xu|467d|W~J`5F0(<>
z@=#Cm(`)nGb%o^2bh)+HAE!jrs4mDMp%l!f%7Azmms~5$0c$vg{^TE1@gFIf+ktbE
zTl3DYMa)yU)Gu%4U4LZaL%fuM6vTY^ARDtg6EjG}&B%gCX=+iADv_|z1=&ls_1<#W
z3gTmr*$|I>@mIcGu)EhHKG_xc;=Z!xf#UQg-U7qf_z%|eaW(eSEJ;3KeAcO43D11P
z2;F2hT>gfp4xpm&ARRErij?b?*kYY4>d&ax!Y+BQjWDE5)_?wq@eC|~!@Ab}>**oQ
zSOc|0pX9@>XJk9HKaU%JTMWm!M>U`CwIZq)cd^FKeE~jpAWH(mGF`tRWI)56baB~=
zH#B3
z9-U`#cKmhx@jO+Wo(N%deSbo6mamsTFBJt9zY!Gc94gzHp2y~0tyS@j%A89G7hN!L
z%HvTvTunV2Nr>0&Y-n@umGqf
zx!ehd*ezi74T7Wnvc;J~MelLJDSqToha0~wvW>C1*r!x?)LK_lhWT+{u!qYsrI(GP
z!n^%D^kVdHwDRo4ex)zY*ZQpY1Bi_`)M}KTz81#%w4_pPfOyu6IH{1WB9|@UP=n
zsP?s^Ff&POA8bG(W1+ORr-d?id}+J2&(1*yi3XpE>ru{%k|YhLCB2XjJyJp1J9}aN
zCP$lmvrAf)R{vrzAr?A+hptRp7Q7YNj)*PCDu24JmSXCPUK~>ZLQq;{esYPf%nOw(
zOHkn{Lxr71c&gQ;)>4_bldp%gYzQFP2$e)|xE+vi{uV5ARNZho!_;kzCyUb!V8gD>
z($9by97NilNt0AH{-9O`X>Rrn7N^?~S57J1w8P<3*ky3lUwh*JJ4}u6uf7pAsq?tbE_etO+I9O%`|O0t2BRjEZA?nQ1SFx
zI0FYL5+&Hkq<%%58bxq^^cJ!Yoh3WOp?}G7`_xDRG`WSrW#i%$)E8jRuJ!YZ$w&>%P3E4r?vhlY
zgPd%)Lq$F3i?ltq3m(22ph)j17r^M!(!fAZw}0uBt3c2vcQj1SbrKW9{v^8?;D2%M
zpSwJ^;34cNq50=t+E-?*5W|Acw+oPt{zGdib!R2=-(WK_S
zyvzS2{Q<4v&`IPyAwbI>y}Rm&w|}A~bi{)h_oz{mL&87k>!pty_$el7AV=^E=fV%TC1yilE`RKg(zdR>
z_VVsaz!p`)m$4RXo}tHz*gb+bWzbH%ha;z@p%E-un~}F+`xrai0^R(&84T*bgX?ND
z*C8RPBJ6WNu*C%p15C#+VEWiz`{I^>$+gs8H!0)a(J$BoEf+CSlEY5tG}Il*-3FWe
zj>)tf`&!A&RR*QTJ1rsrEq}Qynu|mf(TY66kecCuVyzJa{UWD7{JvOVKyS`yLk(gumQEd6NaR4;l1Z;O{(vkfNOe&L
z{~=8V($@1Pjn?L#tUgcJn0m3Z?8W|K;Xv*T$tb;~c(O1z_zXi}Ee0gR~L8kJT}{
z7;oydp_Tt>ym_6RkbkxtArw&`04d*7K?i`n5HV-=iwC7LP`9+e&SBi9|9NQyia%bZ
zgi{W7mG26}60>mpU9{U&C69y9n;kG#y?%z=t!{oqmY|q)Wd%jVaW}LmIM_E
zDWqfh6aL73lsktvZuVE!
z$i<13j|XlPwhG;4&@|jAP6)%Faon5#x7+LA*}Zle``C>+#wx1*MAgxA!|=7Hb$rAG|_^B`h$WYvV@krqVz4J|m~_^$b0~
zFptMeNmJ%%my<5HuZ)vNuW^Fs#cE63CSd_>cX}$|
z569E~7Mqn+FX*L=QFu=PE+!x3R-A$RP5}cAYeIZ#rD)){mDoUt68{eQ&D@S6HE30t
zpGztmfPX~7pQ~oJJ8{WK#&vJveNl*3M8BM2yK%61NC1#QJUw<9wl?Ev_!NB`Db}mj
zkr#RIQ1b0NhIC`Mb_dTNH4lpvr2bxzDFixU;MD!mP)mKj1|QvMgu=c&IH4+YLR)E(
z!j+)(&lNfV-G@m_rXg<^E`8w^AbZimY@1h00AB
zmm5-cy9A3vM@G*=&2*eL29qE;i=lqovS6!_6%gD~I13Iw^9PHtE$;#nL9iGHMG57g
z9>N*S3?#V-VFj}Bp#462DAd!JI#Y1R?;{j?@d!S&x~leyneH@!*U~^`BV8Xj4nm^?
z+J8{W7h=oELJ0+D|>g
z8Mlvz;bT>acjiS{P~g|el`>?d->^_4cCN_ku4!f)+pn4o%}`5Va>t7_pr_PYSyN6U
zza=(f@SKF)usi-ap*S1P7kzrR-qz
z!MXGqh7vn1xCaE5wGEY?9p0RSrkrqlHUWy;$Y!fFB3~!4WwlQ&+A`B3#>R-ZcVx!4
zm2Nzbt&VrU`XMov00U}hfzIV>Zs&RZGpd^ru78q{H>_1uU62B|8VvT^07DhPaUn;#zqaPYxcX3K
ziW3&@S{w%={Wb6&{k()vWSO-4JZMC9o6@kPZFB;@ZZGz&>#S=|bh+W>y(V#{wu-Ly
zoYI2eg`%)-^G2--Xmm_n-7f!l{O(>HjTDDMgLO>S{_-0~idHB#zTEQ^j(_mfm08PP
z5iYL`Nch5keZpY1zQX%AMcgdjra@H&NMt{K(iDKsvI`>YjO$rYPWyF!;A9$;mjOyC
z2KCVhg6YML@AM-@L>ra6y1729SFc+#N*|mgOU~Tc^2qq7MipL7TS2sH7V2^yI$UB@
ztyC=;L)w#;s7_-+SRa}R*MHwyH6;SbqbOJ`hggyrbD0&YI$3K=(`r+JD}*$+Cs^ik
zJBbo!?J2O&4E&A1&VJCXA0pHLf5h8l;&!B10algj5|&T9pK)G
z%Q8BV6wOeemp+k0uLBvSlu=%(HR=J4CmTgujajQxbIa)a4m!Jf=YO8uu&(k4&8R^P
z!(*0|eSyma$Pbk$RX;Nqr?z~)>uhwGB3fmZgNQCQ^cy&K%2-z7=Gma6grJP9BJ!5Q
zuhb)bgv`98#uSj1dePV?Yze>C2SzhHaZO^J
z_8xou&yw^I$fVZJqkmgIEq>{auRx(NzG%s$gJTA15m}5pd>?%7HI2qtaX`}UdD@@K9n59*2exDi)jM6QS$1h$
zmRc0c@;AzgjB#5jt7
zjii(?Ep!Ox`n6E3I4+lm7}!LZw^E>}ePfz=+xPs)clp?`r5t*?i{e};rI}g;vFWLo
z`=@%QuZAi4ae#q?KS*vus(ESe-DkPYHr{`pi+4U7z<-^VL_Bo53$`iH-RAB#SBws&
zz8FvwUmrcXFjWiB;P-dmALQSW>&psMH`M)FZTZ__NQ?a}j{+B_>T3@8w+_KXI7|5_
z97UeN-@Pe=K)I}QQho6fw2(pQQWYXt`-D*2io#kE*8PJ;z6%1E2h=&W*b
zC0G{y-F9eWtZj>hfgg)4S`lYl)n4qM$ay$E?|cZ>Gk0Sk#mnwaN+5TywOBAUJL6+Y
z1u6?71jdDWHSFxFv6PmXU*8X0!&e~~LJ^|8QGdxa#|88D!%J?^0LPu+f6eUr;~2m1
zH^KN>ZcDbSA*t`sAqSD(>Lcps)*+TBNxg^ZG1xwI>6Cf$v;(jVB90A0RBEkEZ1css
z(xHh>1hN0DcB!2E9?WayNmm(4L>`d6Dv&9J`UPqfGmv#|~
zo53q26-Y?>mH>Y{AtP}V;p!nOt$ampXOt_ZsuUfAWh>d+`ql3Bi~QWJuCO_EwLSHw
zcn$=SUhrYv*wd=DFce$w34a|Z??bOt
z=<8I*X_eT?!mtq7tFhQ-ypHj9V_SC~1}T_P9=6Nwe=mf*{#IZwcR0)CwWU)lL5UIE
zWH2g_G?73R`AZ+rJW|LNQ
zCogvW2{g)q%RaaHum*wM9)HtD#Z9a*j;BK!5szH#nx7xtOm;FzWFT35V8ch27$xYM
zaWvV9KEc_t2>v|zGNAEuAaWU4Yi@m3JJrE@eek}Bwt&xMBw9C1qc`EJqu|)9e2*6F
zWr5+ZR*{E$a@j|KFXAEdLx}W=+K8$CC(Sn=?Edd>^>(nzgjOwqhkt7#2A~VE=+boI
z1M%skm0ANiAk#t2GXX-@P|R
z+Vx*bNe(wnVVaafe}CXkE2WDI`{i-IKS!y1v--aui$=eNH0*)1gkSs@Sfq9lf4&cK
zESn&>ETp)s$w0-PNNA#*x{af5LI4{keI`z)879*4ZYVEe?bBQr+dxm$HQw6?>(D8b
z(&yxBd)+#i!G=-5grO5skYK559oI^#VWkUbYe8)g?uPl64|2EXHCUZgw7<_EYlkLI3h|3
z@g{r2V~l(fs-6D2!*j8X-fO@B!hbW8$CX-!gO2v&kn#D$BcmBy=pRPTIDd)VL8RWg
zKJZ+hMlnxhK7W-=EGcklLYmhskv;9df(pgAd2b~~Y`TpMuv8(6*c2ZSJMnU1Y!|tR
ziwlVapTO4elanBg1*Jc(ckSH4>~dErwZUsSK!@&9^P7NyfxxP(?5e~2GL_8Agycy|
z$1QyOoyE{CVc>@$P*#;&mPbGR>D%QHb
z#FKD|uV0lqnnZ^@RSGmhjIK1?OBF~_6NP9oUt^-VeS)~?l^JR?FcCswN&ep)PqaC
zvwvA7aMBzOgRIZR=92Ipfn+S%6Qj3~dN3B1gB6lW((N{HI9zAc=Y;jM0MaF>1^lP0naoJ
z`H3hTGbOkOHo^?#+*LrUkc0g*^zj99a(`w+od9*Zqw20(t5_y1eg!I
z4ER+Uooy{GuYoc1bL&7^ub|n%xlk$8Le)(;LsYEisII#QaF_V%BRVKQV)1_QttA0*
z*^}FHUgpQJ${jk<#R79_E*k=X4F3sH>RS#lAHmIZOd$J#kcXnM#+A^t7<*xLvwutP
zO!r~E*E}%4&J?0bS4tcL4!gohqgiE2$rPzOW-H;QQEtvq0if|$EUK9PKs3TSs1GPn8m_rNH!PPwddxA9e9QKj9ig>^CI5z%9
z+|+YY1O1Bb@-E%jWe$v&ch(^h>^>B+v5FoZSe~@`d6*q&KyfGi#3pC3Ab+S12neSe
z9Ie`XjIMC5Hc+*Fe05n06b<}!!b_ff<+SEQKp)^}mL7oM==I<6o{%3eO7vwj^14
zd)UogFvBgUbGB?QTPZA>Q-9?|8Q`}M6_qsZ-#j#Byb6xs|MXoK@`So((I_zrx*LRC
z`+N4R@1lFAdrpeiA}XE~UCsNO!Fm*)W~FcCnoACQ$e8)g%a}7r?_-fvST$i;Ny=(u
zj9MAZSa}5(Hm=RK&GgW(Cb64|u2p$(EdSABvTL>&zm=HR0r|PRXn#hNERALAXIZo4
zGzw8nDZ`MA9h_u>h41M58`<(cWl-4Iw&Y|h2QgVV1=WkH(ReU~W|SY;23nKbD?hM1UEg8pp_)RlmFbQO#Puj{Z^nBXJDeNFGF<<;%qrC6tsC
z`dkj3!tkfj##DXBXn(MUG`yg#*T`zyuY2RXwep3psjg$A-DgZ>#*A6vb1!IqT@C>@
zqbmDS4{Ozyv+U%)zLX&M16qe&mQ@W$2M7FTJJm8y1MWAgZ@qaKldn#L
z*DBWOR^)=?84jl6kR$VHUsKYvqavP4iEg}szC0*d-Y%vAwTlaFlV6|Y0nv^FpAYDQ
zJlLUdpY9HaGk?d61UkS>>~?2u`6pM#nT{T9%K4yHy1inN>;V-g8rbQXQC3`&}%Jb%1oH}lS2O%7LH;d9?`hLY@;eqBY?*m3h0
z2?BiAl;)J4R>1pTNd)OJWf5-?eR%2y=Zz*g_6QKrw9}QCoE+%;!XDzDKgPI=nGGt%
zTOSI6lxijmDwh4*Kc6W+lRWXpsLhTs`|N-;2|Y+=$^-&H96zTw_9|P>1>(8mu77~B
zuW`rn6d+m$>pyNKWF1n_S4;o;&1ciGxPHbQzwG3+6e)G8=hEtKr?V^1pU%;n{CmKY
zB(}OaUP6KH@tP-S@X`n#%X(5*{zkget>+jk`TjwWo3JX>=Z0?PYl)s1g5y}sras30a
zbU%ugMPL95qza=3Wm%D|7c;e@4Krw2E{=Rk)nc>E1#B*Mk&}>{6qyY~hfen$WR#h+
z)((etHGr0@dI81XvC_#6bH|i)MZKX;jj0li
zqjIiY*R~UQ?^(zN4o3D`tK!Os14`nF
zA?>9e(SOTekX02rD)zWb;%}h3I2Hx$1A2D7rLSI%KJ8XyCyohBji#C(_n2$dSc24H
zmt)z)?tct$CM~EW6Dndo
ze=ux<`LQEZ83h$tbY^47RaUIf6ldLP!*>BU`+r3ifY^q`w_|P@?~J@r_=ILeCo#O4
zeonf92z!>iUnx&ag3$K#fR3GzSsvGeG16X(-#|{S&(T%K&OfV+7>PeR+_&zKPw&2M
zviQ?I;_eDgOURu*>wnQER~u|kA_g683^?};+tTI-a8Jh_u?KbX0K*J2nw%JnPOq+~7I57PR%WaXmg3mo~t;nqy4okZ%wgoV#>vLhu)M}>&s050xS=q}d
z7>xhEz_5?hWS31K9W(TZws@BAC{XPwg7sLVun@H8up&!6$LH;jCCdt_n*Y_ug-@gs
z1ab`RQM=Kv6@L)fz*WKUrY-9)9x%MSBHkRN&ZKCINY?E6?4h@i1;ci1!?tQRYDPyg
z8gX9s_i$yehO1zX!?gny
zrhdGfTz^+~r8I;Un{Xy^>Buy=-gj*XsT8s_ATs+G^;(Pw-NcmamnI6lz+Zhv*6~7l
zg(nkDnoQ$DAdo`_^}aaaIk&+kiER$CDiRdScZyLFxfm~d*1nYNyO=cuBoPIc^Rbe(SGWr(=
z=YP#&+iGv{T+8}Av1UE+_pQEeEE@)>Lde0`g0Ni+BFk~Jx|{MEmWdWkVKOQ~LvVM`9DkwTNP+X;|E(5^zf
zZ1bMb+ntYTPbL6^Ji6W^>hQc=7ft5#{`F#xeODjM=XO#S#N@U|`2;|U|mqn<<+U61=c8b((|N3iNKFvILaYn<4~0Ks)CW5q>wgUZ_)QK153csS2-|SE3X3y37q@tQ*!Jp`LTPS(qR;I(8;_
zrcikm94x$xm#?+D5mS>Y6oteLjeo3!E0X~1bt9#P|#1GzSK#iA36_Rf_f8Xpacs*gL&D7g2>PC$`cW2dz=?uhe^L_X($7
zz5?>@gH|IV4?Z-cfN|IC-y}0k$^dbrk8{N`{#qD?W+op*^Io^M!$XVO2!Gb4gbuQm
zOjYNYo2nGBU!4rkgAn4BU}~84%d7)n>d@L5H4IT)B&E>4atL8fhUO83uZ>X?$lpdJ%uQ79_1!9VuZ8hq2GfXgYj7`HwD_|6^9Wz1-(mx)S@iBj#zL
zRrA@*h1kmgr2J!>l)jw*TIU
z+-Oo}+^oDs2lB?Rv|4#hb~@tdgLQWYdzSkZB;mp1*s34XhaAet?AE&`Z1(`N?z?=}
z7^StmQBagy2GXUB;G1>gyPo>Y$uUot0nt0p_lf@r(-{@O7sGo!csN;yg#={=g4#OO
z`_{E@x3>p_dqCLy?SHk`ASOK6O;lv$pH7Fvj3npZMHn{pI`vX!xH#5wSgUkm)zlxL
zJG4-f)#g@0SZ2KhHLio?kc!7cLw)i$+K9hMi%0$fpFrI*WnGLJflL2I+>Txc+OMzc
zcE%?kfKS5NybO>GUUfO5nnu=R`q_@Q=PiPrkm{*VTSyg|zJC|u0R!XkQ!=Z67>v}_
z@Ho!dnme7Gy3hy)@D!gwAIf<4A{ygKdN>3a?yW@A(seqT7oR5)Ql=Afh~q99EG`w!
z$-@+77ajjJp++Pil#)(@3{J#=RKFiZ))Lx5_0U~p8?_y_t!Xf3kJFa?E4kw^NBjw)
zJ2*%DN&0iipMMvk42{LKD5Hn99X3_^RLFuQ8UP(lINcKSyAjA0pSN%g=^k_VLZ$I5
zn8U$BUECQy3Ts))@@H&&ekj22J>=VxB{QY+O36eBMke=)D(-OQp(BiQ=dMj#KvS&wHAZh!r>w}nZmF@;?_D@`nV&!P}m
zQQggZElNU(7N@^vwD&|iei`ybJsW`!y)fE{^bcUc`f4u(>-0$akOhR>a-S|K@0n~i
zv-`YdZ}pt5VVT2mkE?R&RH_Bf?xEzON*vc)HqeAmUS!l&L@+TkJ1{uvuxg2@m^Q^|ETv90!%+13J`{z}n^W}%ORkc}>%W{^pnwyMK3HFr1kA=6s9ogQ|mA~Dw>WxH7?)G`!Y~@`PqsT9PeoN)ax83rst|w!;
z`%g`#IkDw@H&YHmQ{+$SbVQSQk?i1~QQF93M2Lv+$yGjtwNKYSW&MQ`>pTPAVn<#}
z{(r>54~zQllKKWDlu*?ZE33^t=#!kPa506ywq;56zlE|+xCMLADNw9Gf>yhdw#Q-m
z^wvven-a?0BQuJx)c@iRY$AB@blV@kEh6l$+r+(14!YT*JKL~iYkxKg7>~1+>z+8V
zgV4`*8?7!BzOhx<37m*bV^|PgQ_q(*D1T^*jw{i2$5ljz6OyzO;&=l5wkTLpH0(je!!79m}eTXE7&(&h1z6E
ztRT4XdViQw5y;b?H>J`j^tB&c{Hx`!FhgKj*2pf$oxGYdW$LOqwVeMR!lj&mHTY{9
zk4lA-NRWb*K`itn39`Dh_sSi@iz%TH32&){lKsh74c8>HurJ1rENLFe%zwVWJl}S_
zGK38|HiPT!w3qxqP3AtDs-E#9sky^J9iPXzCPIkV6X`z3HR0mVgRgH@2QJfrF2-c7
zn4{N;YscvN^?p}!w8!z<{vG9!sh}wa;MAEZTSAOdYZNA#wn--{o+g2$$J}Yh1|%4L
zY#oM>vDwz@$!`!5JNu`EUVqo{QP(4trwmH#Xyfn;}OjUl8v>uD4KTD4pT1^Z|Uj5uWRWk*b^@4fTry9
z&(&wWAR&8a-KKmyAQAd#DVsXU=|~oOJ>ltKX2{(dT(Up?=h^V(Sbs0|+01j=75ZB;
zya+v|Xr*<^vW$T3DQodu1RejBB!CjtAUwUK~qYTJ@NSOo;U3WhS93`+sFu7fO-yqR1%-dC{8v
zlklC_;=Sg^gDHW@j@C0WbYum?`jT%4?F#Rz4@;%`hIeysnI!_G!o-qLb~m_q0aLg<
zeU<#JnN!U4k!)8Tz@6H(fcj_5S3DE-e1hvSsyj6*iuBYJASLOVk3>fdlgx}VypB2s
zd7y<46CVx`;p7Zh|m>+m;CA^23JhQ=QdqJ!sC+=ETji*B~>6?Un<{bu#dd*G>
zL)6XRNM1+s-37w9n_RD3CPXf{A|(B9GB2Eqvz{XbCNaPpvQ3-bNuQ}U8N=ysgSO8z
z<1;COYuGwmqKjFDeN|B1!IL%aaB+uR+=A=HJ-Cw_T!Ita^@jxr7Tn$4-GW1q;I6?P
z0tAOF|JtqZ;oFCO>oe6gGhOp^rn=8DD-MXDTXW*E`*qCDVmC*nSV)+0S_ce3u3G%0
zkq%DFelO!ilOR!+)*3&%=mlr&k4t4_wSG7H2-66?ke>MAe`Libf|3w%D?0LRtaa&8
z$a-n}18IhJ+OOJ#-HzzTg!)-&mn5<`sW_VJsZ7`@)P0BO0orS29S5|7`Q9R2yf$59
z$JMrog+m>%TpBm+@@Dg7
zT7tAE1{fi1-t#otv3Mn)8!pVT;V@BEJKNI#IjF8aEA+vMu6~Y)rrEY=%mf8qT;{A4
zpSDnIcg=+V5O2Y%Mzpe#lKb24-&tp*9~vY`EsZ@BcRFwVgiH?w1TMmh|)Orc9Y
z9!uw*(>(|{T>Rb5UdGJ|_-(ArVDvHs`uXE4*Rw-~*UpD|x0k-MR8YpjFzY;(1Nr;n
zs_|8l_yz(EBh5;oVljh;^XoKq=sZ)w?=gCzw|=M4UF$5k75-?sVLE&qNY;Y$vGr0t
z5b58+^JM;NHIWN+L{lG?a12DMovV6DUZLFI;wD-YxEcP4wS$Y=sHLS%x{{K#teR7f
zv}H}RxDs$ne-?e(i@~jF`kR!
zo0T_`HHPa1d)Is)s-j#z{D?h)UzeFG=9ce}sM%_1e>x4YP{W7V99}Ja)}ml*I~fNj
zWEc6YtuwEGs+wdRH(lDi%aDyG5HRpRmP5m*h9gnLpuZ+zsA|I!Z=r2JzIu1?Bj!_C
z#bH561ox;61@6rs@b}9n7nOBIh}039s9YOe-L{1ZQVrfi0Qo4a^S;putG+%@Gktwy
zhW9+b0E7o{1ncRAYviUJOclY&rvSX=@6#)lNKE#d;hEhAMug$g+jSGJN2hl_RAJ$;G|
zX~D{!qdg^eddD{H6l%YAic6MVw=MDuRy6A2H};7<7}1KCm%mI}8ofBp784O>?Z)>F
zsu@YIDm)3!ostt7K8qInfAp8hQBoiy6biidBs#MhR5-jLrP#8PZ!G0MwGAN^6E#!z
zJM
z&j5?JuQl9SS9s*19CScJTrh%$OAqy0I&R?XceW|upg9$Pu9`@J-YKOCAcgl_*+
z#Ww!ZAT&onS_fTRJoKRQ{8T||(`apQwj2ax5aD*ZgE?ioc#*&%}e
z_qT0!*bVrEcuOV6b%B<9PfNuC{#4VR3}|xT9Qwpu4pBvK6|VF)4DmObYEDqLU)^25
z?g$uLPvVTrIKv*6`wSG66|vYKzy}E69V+E4tE8$&x45=IH`}CV2%Y+VFPFLLjkHQn
zuSlPsIpS7&O_+$ZvDEQQG!+Tyd_0ZwqP0b!6A4D`qW+xLImik7jF9zZ93&xgOrz#O
zVi%^WJon2a|4n~PfajoQwDAlbOGTXo4#&3qJAOqptU(ZAK{N*?UNH+6o`tKuCV
zl9~*Huk^#!4{)66N1QZx#^Q@0VeX(y%X}wWWhh%=_BO)oU|X&jmPP@y$p5-Sm1sGz
zz8~4uOK68EM^ZSiAZl_s9|8!VjFCfr^x3MMN<_4pe7Uf_#@yE<9`z;k0Dq5!QBY{v
z&`VPb=RPbX&LHgod-DhX@c$(7ugzcZqVF+#Bt}c%%R;ZUi>croCAxzpJ@?-f6cn
    ^_FUk`@Sc3?4{pTDkdnuyXT#!Tbz7&30Eqdu;@m
zQsxpfk{**|Jzz%;i=T1Bp$DwaFX<5F0Dl62$hNcR)Eb5fXBkWDT}fkBIoQa4zfES!JP
z1DpgycZFRVT>~RLYzBaEXvvs|mQJT7g+s6Ji)5Sk@Vf4O#{>%np_ksQSwYvuaVAws
zV8)YEAqzG!D&C~C>DkW9w6Jngp_IhWvfTj5-4WxkE@hEQu{K*8%n${SLgbsiqdS2n
za;7V6s9%<6m~i)toD=g@y$b56jWuD`^l0*ceJUj)F5*dvgj_U73z$rQQy
zW&5r?>Rp;^Rj>f1&HkD}or92;2*_xN142=%HL0%wW^8abc3417aIwd)QJf*YiI4Yg
zp~1#czm)+(i_~aN14*!}^Dox)`;H(eI4l#-rkjt0bca7ci}p)kuduOuG#0ON?d1OF
zYWkO>uncAFLO??1p<{usf@;L;A3DmU%;K95IKj!mn22yS-l@yjW#jS`Q!3ekGh8
zX0l`8PRc>7e~$fWVLKPZ{vfA0h0d1mnY3|MW25j@mxy!yiDWVNe4eff)v9gVQwXp5
ze6+#$A3LCNLLD6VZo|blxc6bnq%qyZ;EZd+Qu7l;$EhnCY;+*ijKUWHtv5E+eh_H!
zb$>sHrEBu}w5Yk~Pbrrwd%5mT5r$tu!%J*NVXZ4MtCaT6ES?I*iQi7)M})fstnm5A
zv5F~LhTe@`LC2t@REfI_ugdCXyu9xt67$Avf}TK`Hz?H=PJP!5rYPD;MJPFP%lF-<
z0(}ppuqh=DgV)Gp>ks6hk5#^`r7B+{bdV8H@f`a%QtX);cB<>fyrdJptcL99LoEz<
ze{gI~_cNsvD+HwPs}_oMH`zWmgpT&`b*&!PAz?eYFLnA#!ujQMJwxF?$0WSf`0bJP
zFE&8RoOaABSAFWW#MI%tK5iqO(G5|YMhMCVu+xidOt$h-aF?GX3Aw+mfjDMlNy15WdZZU+Z!TP4#*
zcC!{?u*eCI7$<)+puPRCa0$V5r~u5@?NG8feFE_AZZ&6mlyEq~uS$i2fes@YP4Oo0
z&3;Fi6>@_uoh~g=G?92f37wpFaK@+po+a~wHPVS2LZ5DivRFb+f*MVvcB*4qeW&@7T
zao7fAn!2E>RT>U4i>xda5Cft!`n$#aQJUA;6gQi9-iZ!GBA*KujnnVKLRG(q`hT87
z29m=2?+?RDgs=-j5<)0s;=sWe=gB_wpv
zm7VQ;>%QDpdY+&|@tdn?z@v6+E{JqilZvwD&H)bY;@PwKV7H!nGrvzl5lRk(ELHXL
zx2VyQ2!V>P49wdlKIqBXjmk_apN1=b18Y<9CY!yPAn}x_o$N=!J;ZaXqiz$
zS=jV$ALAodL82vx+9ogN*Srm_{1It#8}vU3B@GcD%x9xS)}AqH0X@=X&b4(PPu}((
z9OfD_9B;--a4_oB$wXN~uJOABgxJwnG|-=4vuRWe)!#c2nwVZbp{gV7j(f@Omkpix
zu8wLUnHd>#K&_tgT_^GAlK}|Y%hM)Z<-4+-W+a0`e$GV?bw6*4&r7cLVmi{LWEbHe
z=f9p1?=XR8o-w!~Knq?ae#rN`@*(7KqPIwPB$Z@amgYSddd}A!FJ|hgP*3APWuekVXv|0qHo0oP%B;dc
zjPT=-w~?PaAmk(&%7e_WaMJCFQJXvBqb0G^I8SA8;
zUd^f{RCs=wY%0T7#CkbVBSS?y8bc`Lt{4D?Uik_A&)Yj@Mf#fcnpV`cpgG}JAWIpi
zUGP{`7iq{Tq+@arKh{eX#i@}|?2pe@-~DUPtey>!H1>J0vFWT%5x(FTOtEMbK83wf
zQAPh!9;(t8OO%G=eqB(`MKpoi)qOR0wITlXCIOZ!$;t@l`I^D~sa#C^9`=B6V*2r%
z$!}bftFxlhU>p;eMRx=S#3o#>B3UfXArPs#rMLHAvMj1OL)S+=!q+Urbm8*iRY-1~
zTF<$8TptioN+rV8BZvQ}Pkcc#h|vn8pJ~0L|Rq%*EB&+{Erbk%OrX8vmQo
z;2$`GmZC&hbwOCje*=t3*bzBc&i??smO~;q4og_>{|4^?I2j9A-v0pnme4)8baYt$
zmN(#o3d@(0^#~V+%*6@egmCbsusp$$|39Ge1Xo4^L1UM-c6Mq@-5Xqs+#(0xVvSO8Y(*J!AB3uI7^Rv(Rm<^6g?y
z4c#*L=g%A|Rd6~qO}gtwo3uVWEUD%SwLKyDO882j216|CYn2e_d0)}#gUwPuj&ckq
z_BWaj5s}SspU`R|y7Xa_S;A??KS**Gr7dpRrGmDMfloc*&P)A1#kDz?h_g#-3Qc)+
zEsZhOU#dflb$4coi!nvvJtaw_<)i;LqhOe#LQyPgHv);tP?5Wh@$h%^K`0@hQ{zJ%
zNis}>p1Du-1g!2;pbU~3JYC)Jqbd`^beqG^d1)04YLa;g89WfN62I*sZoC75gDiSvWHWtK*P4_Ohc$2+*JIz%jh!0`O>xgqG
zdq@U_)nHxTt53nlGQ$?ZG!i<)Gke0DeEqy(qGphtYOo)En!)>g-67dnSl)_3o
zKm(RKAbR8eFLX$i7=Q>;aCVQ0>Upkn?mWF5qd%STD>+=w=Iq4(DL6N8>otlB)rM3S
zkK?T)iLsRwWXRWK_C0L(n0RemoJaZBP9i?AYq?AKFZW!bF2pTKCqovW`~;uyHfV*g
z03?^MM?HgK{K35`+KJ2>x+j9v_il1cxD<#ti(u_ucQ~+Ku$JKLv<0h9QQ$|%4tgf|
z?BMBLv$eCN@bK3CQa$ZGvk^1zwI}s1;j(t^Z})kDnB5>e4z5*EbHunax`N3)QEHX9
z8Il%_cR_vMopWa7&-+1N$Nb*;@!VlP091de*u>JKycx^~MWu$;{f1P>-VtQ5_*Epx
zwH(sT2>(*vee|bxaKXnB96f2WiYA~i+4XL0KNM&oWU5EjHy6<@?2Gwx4Dl?nP<&e2I{ZXH~apv+Wq&cMfF!kLPp{twbP%oefEAXXLeIB6-)fv6Vt8pFK
zc%5TtIE85E=N|Odiw(5N`s2=L@6TA)r*ttP0hTqgu)x6$LDq5rCuQp&dTOr}Y#coq
zK80dQW=xQ>7BT~UjgWw;SAMQlAFwPv8=lhr`LK4m&S_zWq!LfpZjGK#cPXYna57k>
zf0)k3d6yC0ae85@6*Na(L~1FKR}0z?8F3Y$KvET&^{QeFM>d0Z%M*0vYD%4zoof9{
z5h0XH-bw59El~hNiWwftG1bmwjhXYN5n5Q5Db;cQDPa1zEsq
z%B*&_Ri-ODa92J_dbUuN2hg>;V$38f)ow7zCqP$#GNYI#=@)^iG_7a@HBSoAi<;dr
z_wbiZMbo`m7oSqrogrkjmscpiLR)WQmAGR1RyfqqT%@Q@G#CTfkW;}CjVm*5T%q2(
z2}9c};|9wsmEXjsr@b0uug@^&;Y|%#(tWJ+n9{KaI4N9Y8geGMfD$Vb#RW$*%~U?O
zd^wAcR9A{f*5mYu8e&aOmS|3@zw>zzrH>h6vuEf<%iO32V$1cS-^_}gcKfk#d*8>x
z?g)&p3RrR?XQ38TUfJGzX>?kf{J
zi;-~8G>XH0)~`UB1D0Iyms*GB8b}ODgJA^4p4vX~nx`)!Do9Ge*5n4*j?=zfrJ2Lf
zxob%yPK^qwyROi?U@&!oR?jB7;6k&MgSim%jncHKx}3Zo%-1PDr&mpLij~AMMiPly
ze2L)vUa8n7B9n4FwTPi*4o9$s?=_tSe+-+ZB^^8nqW;4P%w;k3&E09$r)uv$WHXu`
zmPCIIhm9)=uVP^on-e@R91V>#<-_IDOsG}y6L>!_gHt82%D02_!S45!3rAw!RXc=r
z!uEr=E#9hCnm1dXw8$erH#)k8hKyI4<*DZCIA`ixfuK%Td2WtNBidk$Xl&6Ibj`+pU
z4f$o`_R?9c-ZVHzYws^INAGw^Cki5*aCJLgJK}Sn#{(rW2prTrz>4iUJfL1&)pQO|
zdi`X700@3OLb{p5Iw${a|J#B0p9u3E81_|y7D{-j7j1-mx<<{pirBooct-wb$!*=V
z8AiLn>!5KL*yQ7Qb!=>jmE4K`{)6!}G
zzcKIr>_E&!)^%v#!~fv&Ej1T0H~0tSyt_
zy}gNCp*9MBI#@m7t3%H%>-Hr$n`5@Q=8*x2KGYZhP-~n;&EU$vziTS0cO>fxA&|v6H(>$G6
z$y}SU#52^?Qd6OH>9vb6YF}h#XPp9>NA9Yp(-#VQH!gW^ohGPdIh7gc9w&SaUljY1
z=@$Dz4ja#J+~AcGGkSvb_)=-X^V&~kG((?LP3U&@HXKzb>Iu6(A3iu}j
zy#tL7TRzpq#Ox1eI)XlV?_;e@=@>pT1QSFE^{&{_mf_Ks(v6t*=~Sz06#>m^UTT#8
z9DO^lkG|^{srVZL6qv<#<}M^lA8WNr5e(GG{6otBwn`R3gRlQua4@p?`*D{(?nvm(
zBk#4JHOxW*%0}%j28+@!h*BxAfa)f4$ibAW+xX36!ba|C^pF^Xw&H}|%*IA7zu*Ok
zz0UUEcD}Sz-aC+}Vp;`J0vlnqexj9r&B3fsDAp{G&7pTkyn)^{?KOo}BTV-m^k%(4WlK_~5GP)_xt&(l>A{K9
z%ph7rnP;Z)%BCr;g^b{=uTBiH=fT7xv*-+X&ACK~b;kk1auHS~UXUWceev<>pGjzkjW1T|v}iQ9N%F5+$GWhz;dSdt58Dhlf-!SfWu+lalV
z`Np)PXF&v$;HBdNNj*093nZo1Hhp^GVEu}Ut04VHN~s7^_CFZT>mCjUwfnd>uJ^Wb
z(7DnhnWMS>qK*UmHt-s*YcUhK$at*gBzU!pAp)_V4OtQoq6rT$5Pik_&T%4Q8GyK)
z?3}Kq{9B7P(+BBycC4^&+rmPfadLdRmvmHhCoNlelRlho%tz$Fq)G6?@zZ(`L1f3E
zjTu%~=%*W`;Y7pVt%2Ah5)O|Cv&)Kh4}L|@`wHHZriOM$_-La>?jO3U63QdFDZ5$}
zLViT`Hqh|tSP72;%JhCaI(9RwC4U<>3R~oNt{YFaU#2g`1|;trnZyo_?kc{%q$esm
z&$|rLcWt%zkajkGA`gkxWyO-zs3Jm^*KZK8;yY4aLiEyeZVZP;Vn7C{iLs^FY+#vo
z^Kr2yS!b4&y;)jC_W6zoerxxQP7og1sK3y6nXXw-@E!R9JKB{m7Ll&M1ii9Vy~7G6
z&UPPV1WtRU2i3~
zAMRRu_AB}apYKrE{J*IdgdfZS{txF87Di*&@^Ul>vwzYsw?Jc;w>L8fJi&S#V2FSL
z8oQ>oxA|Ly#tsGRae}$QoNsFy4i2th2=Chl8vn_+Z~#O2{s%~d^@OFRrDZuKIV2^d
zBza|Ic_n4J1UMiNK8U0gH$S%oCyyxj|4PB;5q&%6e@Gb@4=4DIlraJn<855QPWYjJ
zu3u^H<6|$EgR!Yizp>uZ-DZvDPM7iOa@%m*OCC|W+V_=~4GNb$Ot?hcrKa%)={Tc=
z=D`$Cwel-WF*j7^z~a^-W;kN_I3l<$h0@DOayRy1k@ukPg5zbhft(;IIVq$SW1RGlmB0CJqp;>_@5WW&9$JGh|FKl!t()7AjxbgReBE
z(DLdpdwcVnm*q*`N4rRC&{ieFp>w||rG7E!=HaxdH1@}}u7;=Ybvcvo2g5P69~Xu!
wj%b#SSVWGfP{*oEC^sEf%`J!mi8A2-d$g1uG0W81cE+cv(~wr$(CZQHgp|Cx2>V!ChoT~FV4Rdt1C!)+wP
z#gTwAF%YEQ2Y`?Ps@0{OHrbJSPSwuDRWz7^?>L+$+!L}hTo}v4Uz^rJW%|*`yxj#B
zVqRBoSGWDBDIAX(k10XTfuenVx0M?@^hhfKT@t?ZKaP(l58bL%Dnptk=NBio#S7fD
zl8qCM)Uqz^9qTnexZj$y&z?pR)Sg^R+mB1K4Hco)(!5K6>5Ve;Rd`jX)fQJFuC3pu
zpV{eKs^nqH)yimELnV)#+$fNw`b0muWEl
z=b~+@>4*mBLw3fTLTyFcgo`YttukgM%0ZEewemk4-YDA^P+eM?r%gTeE7smak4c<<
z0=;7gm9zwaO_6g}fPeU-z0SMQaz8_0dH|63-YJu}v-JZv&YUB`q#7nXitYmQuoT2Z
zMcdr@yp~1-Wxmpkt&~x6`SQq+iK0jDdg#V9{l2E!JlwD%aBJt`cq}Ai8kNW$V^e_pJ0BzDk?UI@mrmUZ0yQe*ZexOah~Q5x!`(5C|=TF
zPJ$QvG~2NeT-19ywM9}dK{XcCC+CyG*YY9(#;3>?4h#zYlP6<&Fi@6E(a4_Pbw+`0
zV1Hl78bEsmDTq!4=A5sB2P|sKwJF!I`XB>vonGTv*qW&KV53ETt1#z+HgRUubT}fY
z?78L~iZ>H8CAi40J{|67j}mZ5Fc$E2uNj+6rjX78V%PZLGPD^RBJo7LHfHJUKy9~J
zhF{c1h>kWmV~*yHX`CC)_exoJJsKlPy?fTVH1qjhwh}csSQl*cB3!d;Z0!DHP3{J8
z_&67IMXV_r%0~!*Pnz!di;D&0nO#|S^Dlw;byn3{zN#
      ;+1#MQN-F_64@Q_*|g
z5$|{}b*Lz*YD&4~Mnven7j=W#71o4tXjuetIdRJ!rl*L*o1}5P;amR@y9J&h&(9Am
zEPRi~r-D3B4^oquUCAWRTW4OG5#W4B
z%dw^?*kDEYImU3dBfFLOl@Ozvuz@V27}HHx%M^=*Vdj?~AT5x>HMm6IEYoxkD?VzS
zYY+r{!X(P2J^P@$6fSGi8BZR7Lj8EabQQtW+>cwmKI^xzcGqdtKX>cU5}&<}FcYR1
z-HiSSlVIImXIFKf4o@lI@M{2)65E<=iUIMf3wq*Z{5y-wDo^fnJ?*@CzKM}Xt^YwY
za%`8;S=@tF>t^?5FNubu&o_Kv#B6eRP-kz$mjD$9?Mx~0bM3AgVZc4$pa{+)N>YXB
zk`9f3WCX(MugQhwB8MLAiy@mO&-YUc&$Cw~Ti)^YiLGu&K7gv&uUvjWld
zQnW)&#uhPabQYKgmQ0vzc%-%^XRP%oH>~RvV=(2FC;ZXjr`9_NNO`SK#e!iRP~X%o
z2++4w&yebFcn(kl(=J)xSLy;K%GBB>B?1Q_7afeJ;*C%dRI)ArE%p-5Vx2}g_FqIK
zbvW-WDq9a~QMejpR2Ht^d}7=%L0-$?AfudFqJ|34)x@jC=YpPOWFp98R_LaM1w!Ad
zoAAG}IKeruIeFp}9X{yu^SttXk8UyYj`R8L7X)PmX_UJCyF=Mxb6zFp{3x^lgxt&*
zD=xb>D{pWJ!4Y%7l0Vo)3OQuC+d|EY&MJyxW<5$!r^%l3{ij^#%P>mVNe!2Dvx#T8
z4Hl`gVBhUm5VsTJ!-S2l6p%_DSW-VMQXPdfNQo5N=T7bRbfUF#5emO01u~
z6K7B8_D9E=hbb7XpVQb8h
z#M9RsT+d(-?VEI%b%^hB{;o(gG>K?C1_P{@@VJ@CL=T-`4NF2h2kLIn4`voydvQ2?
zLfq^;>Fn7@9LBfn+4Pg^a;VV0wyX9NVRbGvqLz@T6*1Y4>h3yQE7~-EM|zc+
z0g(&D1D%|znEn8y%i&WZd;M+iA;vJmQyTZ_>3Yp}jtO*J{7aBm4f;gq%^7%s3GxTm
z=wq>Z?c2&d7H1G?O2FLZW4W{`ym9L_TB}@B`$!&;T}+FlU)O7L(|(f)o_?$DHwr3A
z1Ntqx!1~|rsO@d;wi`~T|4_|Kg?H!JFkYO`9}L+)GmDsfE!IVKPyu>^U!(=^LRp82
z){3FixepiiOvlaIHMbSMKgRdX`T)Ux4-|_+jTreErB_i|4)I+iC<4v80(cL8osF_l
z8j1$T{4`~YJk}*6!&0tz2Cj&wsHrjb=qOzVVYD6Bj!us{v|)Yjbd;EMjl?s{4kemD
z%q8Keh1b}eO%p$Mb$UDkazwITN0)}Be}v(T?d7Lh9e
zJKATlBMf1K=72#p2;{}DEoxU5$KV~Ne?ONSqAn+hZoimh6p5?W=jbg9caG{beJ&y(
z{{H;@gyQoJt{PlPW45P?M-@D
zcQ1A&XKz<3ye0)g@pQ49IZ&XXY~+*D!R49?KjxXS!;hy8W%~}T6l9V?3K=xB0GlJn6Sf<^_HDKr3RbVNr4
z2NS`3tyetch`Yli<{^`85fFTWbH7B>{mMI{K5t0uBe95*Phd=h&>zQi{D@uwv0{Uk
zfXw{+wZ-=|m7B1B0Zx^L_rUHAfw4?C@YRx7lvn9hWRl=`;zw4xP%5#7%
zV|^G^o~eX9>&yG4xzrc_Uh_Cx`Scy$hfyU4=a+Q&=jvUk5&l{54u>=~!U+8mU*bDma~<$!oqQ(<2qotPfq7b_$9g${P)Jp9ujLm#5&$u5A~2k3VvGP
zYA?hoG8ecnwiVi5cPF$hAJB12jaJujr463N^Wvnv^FZ)&g9}mlTi+TNX5WO|;rZfZ
z=1Gs43k?T2ltz>Dm{mg9RAq0Q!NCBGk;PLMB`3gb-pcOO@B^E$~2QT)i>^`Awc>R(3T5km?+5BpoCWoo24(jm$uhl%*TZ!SPlX+Xj4XyPJ`
z^xDbx7z12QL>3&lK6r-KjwU0|A|kzCW{0bF#jpdaUm)+sh{(4$FxZZC)P;+p3Zn24
zvo1Xhga&Bk5V>&r0jwh1S9k(NRbov8rdw6%Qe%>gZW4HCOQ9WITRU#N3@SLlmg5`5
zpNt3@avL|hUfMtQ?f?Lds+yUIP^b%O^_6_&kR~jJOY&1w!}jd2I;RqmLo9SS7@XhV
z&P3Xve7}h_^taT>X0H9lJXkk{zH>Zl8!S8B&0{H*KhEnc0-gw0Pri9N(_j%Oks1Mc
z_yQP_7YkkoC5Awa$O$Fy>}C3|Xbyd(_WOixB6FbhOv8tl5*Fa$aZ*a(kveWFu@xMn
zcEG0@3LbM*8*3KSPrTLp+?FVvL1InSzRzKOZ0;aBEs-2tKV-N${!JU+!L^boH-(6d
z?13XoKFii!Zn9p@Zg__r+|sII6LHmvl8`=3TX(;StO9{nZXwN4`NR=roq$@?)HNkz
z32dXa&Itrc!~y`JD@F#F@S7A8j27o7ANz+ADFW7**ey+{QDNUt<~&3l>?jH4?{R>s
z3${Qw+{=ljhp8#k81AO@3@hk;*C+vq?A`j3F6U6CkPJO%2RgN0$GIsZyiz+Y4R=PC
zmr8@O_@ww#rHiN>FsM?@d5%2GRXDhwo8WjXsV~RE#3De~z{e2wq20m`$v)u9>r+1S
z&WoTL)&WQp`l&Ck6h~f<0)p*I*m6uuvZsc0w@p!!K3=>71^D*L!jyp|r%Tg3x)lb>
z-?**r1vBRB5rMg8iWKH4#RPg2fQfq%h}-;qaKV_N5Y4Lb#>aVkl}|b8GfH~NBO7ZSTx5KDbdDC+EtwgddaJ?bEhCaNT{rg
z4tHso5O_CaI#||AI>(pCt$kvxCJs=QX1?xY0ZC`9+4NV9K&)4QRY?j~tn%IW0f+7`
zr8xHD4w38IHR74@kHk(MkqGJ1M(j+T-Ko+D=s1{yps=_FS`fd^b9@hSJWUP?)
zrUHaE6+}bG7asah8Oc2exnU0IQl;ri!1ismioV^2teb4p2_JjsiSJj(bw(Ja!7l?H
zbFy3(m67bnz#rZ7!GZKb>r1lDtg>1uZ7Ioxz0BqGf8OdsRy{Wc&w%c81~Un4A?L|aTN{lDc}^1NuX
zJ7va*2u8}BhsPc6g_=Pm-0h|
zAG_TaOk_uP%_R&%EuQ(2&^gw@K})HQp|U~IOvPVP{RwTA1nhCV1!PtC@2n@GI{=+Y
z6EihIJy!?zPa$%(aAI$HQN*&$@O-UHBx&(fIW~OrYl8GdlLOMlwbW7*cA>Ot3>AAu
zo9X4X5dGWuci9N+V}GG=ZaV)$??Qr+stCgqRM@Sxo*4x
zXPRF~4tSJ=X3?&=kg+8UxTJbnAfDWPSzDS;yRg2{j4g1`-(s(LKG!@MQn=H%0!<@w
z3jubo8HS7(4<+taa1v2a3oY>J7rCt4S!9M%Y}*lM}FDqbfpdN>zU$
zfAC{y>zI1@{Z^-GSjLT^On@|!4IfSTi7L-F5_mw#%{qNDyrdaWF~{{_D@#U*jT3XW
z`~yBig6(iEd*mQ>LZs=R6qUhxT+FvA7vgVF?UNEV{qd^AJ===a45>bB$VS_{9BU)F
zD|Xv`N#!`DF`w&&`hcYPvuH{@eg#PL5yes2qAgLG`N4Tavu8r;0D!}yA7fZ-qWjgt
zuQ2(ElSGF-hVY_3)9*&%WNZ^^R(?&*LpFcQUAAvZ(#Xq2bq8+<2vw|VXzPWf$obnR
z5|mz?!TN8fL~@p!_=t!*=55=hWvKx7_Sw;q6_Ppy
z9_*?;xcUWW`j+mAEZ|=YL*f|4;;{g`gaJPMLM&5@!;9T4jgADWh^2oeY}jESwWle43Vfz5v>2xB`#OBkj^I%rf-c8>qm
z#K{^GPMc$>J+rlYH1Z-B0&YgxS_$<^Gvp4IRT)xQk^;g?!eAgcKuuL&AKtaQR`!S#
z>*%Z>&o3|0_Xh@ubUzVXN?d&Q0cI5we`Gh7m0zcm0d3NDXzY}4(@nRU%n>Fw+y!fa#+k{(
zQ_%2NH8x?~L14Q?&Y-L9zIbX$tl(VlDDLOc`=
zqz31Q#2b`x6XoKP`wO+LbWKEozN+t8O5mm+ZDnD(rGrY2=
zjW!uRZu|ki2l-C_RJCscwV+kr?-wRIm1`kBKc>t>wh*0C5MVm@>{3cvz@Nk4p9*?+
z%MF+1qYbl+_++$m-MpM+-Mo}{G5p*^<{v*y(djtM#Q4+82-${xCgG{ROWu;^!{L@)
zL*NY6UuT!ZjD#m9vhS`SK5hS&0e~xho6S}Wi9jaV0BIjQm;hH7!QDK)4A1S3KCJWE
z$%QDWkiE0!!ovnMu66nvpv|!co&b*ziYtMbK_8fR*a*LoiF1H=SrUD~$MbWCPhHeF
z*a6EUSAZBnsnG!@BK_>lRz^~)kN!rqbtB&RK-d|quLTP)3uk9B7B>$-vTjMtJWqAK
z_E=#CDbTX{lFm2mspnp>lrn*V$iLexL$(u|RG=WM^VoE|mOr%v2+tx?o)T@y$MU$^
zLIyBzv&ZKmxwF7cUj>OTF0ru}=m}`|TKtR-bQ}fFi2tLArc6v2|7fzt%RE0>Q2|vj
za9#rSx`Ys%#(2pOS{Zt*xD+_)LndcN>w3&DH=`M$x%>rEuz8c+bDg%Dtgzu3@xxQ_
zW4YqAnt~dT5Ky257-WeE^wrXh8(W@Qvt;g`{*2f2kjfuoL2^{r
zz^}S8+V1CDY_6IrfqOc|aQ|?&@^MEkk_O5$vID51$TE~*1Z7&STXt^sCH*%N`R6yZ
zZlNm0T@ZcJM!r*UKfeZ*sy|_R{g7wjtVYxt;IszVoxGv|^I{zOv!V1-`LLEw&LPzM
z$!-a2M0IC}OxV2o&-=i8f%3%X6Q}H~r?%1c<#3sn>+v*2tN+MC5o~wekxouhR7E#Z}s#11viO4X(2M0
z6<$yV?ef3=Y%^;VJP`2-y;n|HGzcEQ2-}h0C7B!n4*5`K$>f5M&@e9ZiE?Ef&eI90
zz~J%s@1n)$gXtU}z&)&Ai_hH-*OLeZ?hB&Rh80T0Nyj!^PAQw8FMbjdo7Iqy46&+Rgw&gq7p$dDR
zsDs1+G+?NLDf23j)=6&xPwRhC&AmyBzXF3SDey({sPG5|n~?=3d`F|s?Ij*RXM!gJ
zB2})X$Tc-tl|FH3*JB#QO9f+G2?vAg%RrJF+1YlH?@T1l<0TC#VI<}6ry5-z>w3!L
zPmxrNpmxvU+7v~xfO~wT7Q^s
z)+ru50!ZZ7daNrQtjWDJLY_uW@|9;5D0uD5xo51jixyT(w^JT@P}JbwuVGXVQjpP|
zJf2=79({{ulcq>082Ju>=f}AMf$BTKetqPjP3!2hCh=mzF3
z4^PngU}XzIA|B@;9w5}gK$AjLz1PqH5k4BrbuS^5Jn-ket#zaA26N(0{7Aayt3H=8
z2$H10vO$IeHyqbwUb+ztNu{)y;T+nGiToDa+i!YhWbC|*YDL#^B9K_~Z~s#40rt-`
z_&`KG;xs~k1PkIF>{`}$P!BS#%)H}l_#~+W7|sT@sm{>Nt%MV#)n{GcQM_0H?=>L2
z_-#uDo|YjFDoP=`Bu-e5CY^1aUc1v-T5(eQZAl-U@sV#NAHuXY
zdeXlFeZ4{C|2*UBT6wHRHWni=|YW*xNS;Pc5&PfE~pW~l$RSu8i_JI>b
z5%s1FbCK$tvLUq;M_9o1Scm%4S>0J;7@(Kq4}Y=GFfP9PJp|4YHl_Ts9CZC!rj5*b
zE(EtaXgQ>U)5sN`{d^b!7w18@@m5s2w_%8#7TTRH&)ry;D&uzdf)}A;9uGMh-r-qo
z%9@vPjZR`<3Ya736%K6fHJHXnmpM7!2jP{1PI0C^`C{W5YuH$(-iTIE_o1eH$}I7y
zABjAmnHEO~x7-+@f6v3n5H7n}Uk9mE8FMKZ$ggQ8goI<5H`N3IQp@EW*D$zBKUxKc
zwJ9P?d^10S0g)!D$8K@62UmeO_TpgYdVfxN*h&+e{+XWkvXhvT?IDmtE{#8zkjIkF
zi(70w4b+k<77VF;~X`sGFyg(XvZI!w!v)g>6VlXvWrKbL(rg-?=W5-89M;ColQ(xsaZY
z5!g;$p>KN*E=MG`bqGW@+!Ao&&*V~>K5!yMLa_&%b3ic<2!5%WRCZM)iMmdz%{xZU
zwA2=XhMbAL!<3wN$CNU}^vwxX-r0fe&aDyr`EBd|xMP^;A|h9e_uMq2OYn#j
z8IVe({_AjHLv^e6YsPc7%o$^(Hyv2d)fWHMX{K4v5(&R;F0lltT_Jkx^tD{N8@y<}
zrW+09^NuQPR1&ho?U%mukTXc4X@PZU!4+eG+Byj>uPR-+QUMhs3+|;*G1}mf
z$r&K=Zrkpo6Ksr(8_XN)S3eWcrCuPIcA|@J1j5-)*8EMFp4k8u3G!ZA`Kdew92
zzSFxQW{U$RaV?a0yo7Kb3{nn#T(81ZbqsSES5@
zp2PnQmbp1YZ`EsseUq8Q>fa^o`3LvQbM-0@TmitIA<8WLFASMdr2CV|KZ4@SVQxZ^
zGo<#kiBM5%(Uz$Y?6BTk4WTb3RUSlIExFZ~d8%$9YCiy|FuAM_@hmnR3SVEfl9c=#
zA)=c=1aJIrd>D86nqD{1@}^ECiOE3~jy~V8Z)o;8j$>BfNl}TRFD}x51o!(BxQYEV
zozJeyqi!|ZDQ}^WVz;V*BsPGAa%>ZzNrYq&`1l)3S%0Z@XZv5yb^+6$^fTg5@Z|`e
zs~L@E=o|nLv+1k!>5d?V1L@i3Pv``_+2&x;QQ6XxI)NFuTN{m&IWwV?04@z^jG)$E
z%^c+#1qBjzPLWMhKKZ_g6Nwv%dIblyA33Ht9M>2jHRHbTZ4)+Pcx|H32n5zFf`-`1
zX@H_KS&Yt*E^8@Z5r3)*_BAY|`3BpObc$Kc(ij4$UM}kgE^AH<^q%I>QwvSlYA+W5
zFc+kv%H-(4=#Q!~$^N}(Ga-DkLdr-nHeX9Z7IC^7L~-duD(jA=sx$b9q4xL(GzG~p
zSKH^ppa|S;Slk0KKI+_Cy{5jN&JHLprkiq|z*HCV6X-jI;~jDm1^MOHo6&VF^BZkg
zAcXme72E@(a(D<@C{
zSgsJuuGW~$n#a>P@*PX%D)IA~I_q3*Y(oc@L0h66l(I87Z$lT{b_@=9R_Kc?YhH#yUU__c>7w92SEll&t?Mr
zbkfq$JdgsHQC8yU9R0ZJ9%5*oQc+Dk8(K9p&*y6JsZK;%U~RkxFU;j8uf$i4xG51S
zl~qR(d6Zbng>TuXo#EC<8cKB>WY_2qG|EA$^0>Q}DCLNHS*;
zm?Fg={(50%le8Xhl`V=d-|7-Yu2Ny-3xL*04?g%D4EGJS108@x1yCanWj){-E^fbE
z&|w!Tw*7y0{{OmpR@O!fQX^9`b^>~W|8e)++%R;~CbnkI<_rXk3~a3bZ(<~1WM*My
zO_hfMMg?QxplAA@nh!ak#Y8zvYlTJ9-fb1;=D%gz4_gY8x`(x;6Tn5%*2x8y3I-)R
zw>Kju+q-@09Q*kigGK6ezGXGrom=2pThUKq3d{;B87NRAZ9P@>1tdc1QWr=O7M(4D
zV{HelZ)R$8EL>EOU=NYevEe6>P*TJ13lv+)!auzAOf!$!Hv^cM2o(p90}k+;v6F@M
zOGoF2KC`SiKZY^}oy+(ESMH~+kv1)%wX-D{n}7f-@{)=N;_~F!RM+VIDevzedMaJW
zwRSM|>nd-@T^k&pfzXF(02*VWVgOC*$Fmc~2BN_LqX*T@+|UB)O97I}qve+plGFhu
zD6Aq%aMWuuuXG{P6eHz(~K+LOp
zKlxoerMZ*qNJuJ4E6J%q!Jf=z0a^pP;WtCif2q~QKZT4hX=3#&1rUa
za6UBFxx2g5YqhtxF$QK=r{P}z9+X{dacu(c^Z>H`U;26>4j~%cu}dpi?)1Qa
ztfBemM{uAVfqx_tAbuy$?^q{Vi`Y6Z0b$*L2A<_)J^{g?+J1e%mPR%g4{9|96%|B%
z{Ywx>FktBcQ@bEKGCJGZK>q;iJ3IWa8305u5R$E|?B6H(kJia|KARsI9h|QgMjP!X7KCVbfX^5S#t8Q=2|3F%)M;N9#3%{GQw6F`7X#+H|L^gG;>{`a><16T*R3Mda5
z=*+@)`k{1LPx9T!^wY=}+rPGzcnwkC!txg6skt7k$A{3fGpz#@xMs$dU+)&c`GZZ^
za1TUnJ*BfN{NpPB_qT!;vH=YA?Z?E|=)rFF*M{#}FEr`qI!Mz`;6)b*Hjk7jl5y=>
z1_0Oe`;GeyR2~@Jy$TrpmJ5TwLr~fHjWxzbx4@Rm3$BQ+
zY=7=8`Jn5}syxp;8|GJi>1)M(pDcPl#aZu#e)%tfe0MSfPFr(t`}r@x5^>`CJ2wp8
zH~pOo`^FmhyZ_1h#%B#o-7$f*(4YzT3m{JI4JhJ2(TBy3L-os!#MZyKvc3o&`zQf+
zrKR;_+ssU&@yh{(gY^x-fpqp11L16I`v9ltn+@?{k57Z^BYqlvpdSF&iGM}A15+me
zh?jBLlCee_lRbYY~2z{Vr~~Zt>R-
z4)o>h{>`xEGt@sm0I#*}2*0pm;qCrTu>7uq9@R=+UORsSuz%aTf%)^u=5QQL5Upwd
z8fLDbsNRsNuqy!z4Uc+j2UAs1&X_inHpPQ9
zjQ}L>T%30SqjhrKhT265*s4uFv!VCoIc8QPt}UZa+M;AvB$_$r-lrd)vAF0q8)Pkl
zd}fM&L%3)WF4s1rYs4HS;`s$y6n%17_q=uv7Fk44#A(i7!AA>PTceJv!q+fLJ7kj;
zA!RKrnM9b}OxlH3r}g#>su`sD)aGT=PhpAVzI;;vU3-3ztP%ZGaRPRb0owHLzDx;7
ztm33udOqgM-_{~#2LHYEb!dU)ZAZdh9*L?G@n_}l@{h~*@CIYB3vXY+4{WF}5E~k6
z40~2c&9fe1<=C}D%dW|JRe{&OILh9#Q?Tgfn;(rVG`l1u1R>OZw&ybZOFMw(!7K+FbE*x&|ZD*bpVm&yg1EDGo-D8xlTJk@K
zMUUrl^t{c!Qi`}}r1vnMamEp=|J^@lTT;1CFw#?3-@cRz|JFSYJt-J8m%H=K>5CgACVfsFL8ghG;bt<$RoQwbia-(V1fKHiQP-tmXkx?|wd5YJ%)F7$F
zTU^~8?S?JgaOoPLrniRnb1?a&&4VYCiR*>*)Qk7~`nO5z>P=TuH6sKF8Ki)XD~xGT
zd)k8sa=m{VxyVIJusolF)yPaRaZU6&YesVl(+H@pSjq=TM!=<20VzpVAV90wJ3Gn+o!aOMvHl3Jtq(f-lIQ5d{P`Drs;=u3hnCptsN?{
zjk>aw!%du!u<&1`)stg2P(zZ*p{c>t^q#RX_IX!R50;$POkvDot4J1kpJQXE+I?ze
zf$5MxI^Il%ZhK8Kh+VHV%bGzBR3=~m6%rB3?7{$vda#u^0W|>!%A3w1A{-lxRS@XN
zumuExCDNNj(Q=&*hNxKy{A3kwEJSxB*j%T_{RJ#oU+x2nxP))}CBD*=YJ16D$OO#A{JQ*H3*op#1CIBOa)lM+DP{RS9gqqg5^i
z&*H&lIH}hRh#Ks0*Zp|(>97YdpM!l{tpAHCipw&HCn0?F6e-^ePAzG)*VEyy+QBfa
zuyaQCjahxXfjw_WPNOEiT}REB0!@ZWLLcZ^MAkihVY%(bsef?QpR@hy{X|VZV$dEh
zV*0%QrTS%Y(!@8b28#hQ!sEVD?*4+U5@k=?LRf_0O>4whJfXAI<#i2!G|0IcM
z3SS)g3)g&!ZV{lg4|3s<_*7+s_g|T55|5fRC#5DjE72$o;yG8@0Uy((_aq2Op5zZW
z%OkEt6wNDNq~#I~)Q*uo15=U@*bb=~p~wEOgXpLqhZRZ4=1Av5w9=N`U+
z<)ZmItlUmQD;`W~-3Cogj1iP-5YgQ*^BpBt$D)XeNS8ORH^6{rU3=MNJi|yg=klq&
zxZ=Ibw3hNNgnSfm@kdYz99OpNT#gqf!Alp3;WCU`tE47KNIarWZWH545LGqV+q0^-
zpch{v596QTT=LSN#tXgTZ%yg>7puX~AnCopolq+>85H+wNW5LhOFyd$IffQNFyO&m
z{iY)_-aeIYB9Z)hWe0b#Zr=!jCf0b68Rt$OF&jm0dq5+={HCCzcaXdxE8r$`;Z=U8
zB)Vv+-h$V_$CJO5UmJ&qe~+YNhh?F=DB4L@4~*nrL!UrLncG{Dp=>PI!Rq7qZVqS1
zBc{f)oEe8pzv<1NKl(fC6aL=ZcW?1(@&Agd!6_C9O3EATsqcyvt5o@JBT#!S#u2<4JS+`~B1lF&nr$VmrWFce(H;
z8AeVxxq(p$TRUmKY}zD7lNh#{WOwX3voeHIzCUxaG`($94TxdT&Ny{Qk
zq9}sSU{)%mVm5>+sW_@Eszx=uKY2moyyt&o{`>Cvs^P_{^#<{e2*aQafOdS3=`+~&
zSB818Gr>0Un5N9l7odf#J@rOK6P?
zI(32!Ssg<+5EcUGxjc8{+o*U%LzNPcHfiJmz1vJd+1SphpiMx;d8n~do78r?Ll{mJ
zU&Eu4F1J4i!-X&9sYd5~v0{FNOvCZuW9o(S7Y5vIOCCcxG5^kX-D$s9U0)l8Z<6RK
zNR=U-yDmP;l$8Go%`BI~*Kw(jbfM9-tw){Jl=3^c1(M~~!O8Ez>QRs-&U^qcKgL`fLSrJgtp
z8$K&tv8$zabq%~rPh*Gth7%e%i{tz~9r)gZYdqnZGk0~+H8m9=_o_&E$?4K7@#@YE
zmq>GFH4;k_rWZucdh5W@MU?=^zWsvg^p1k4PepXBYF@uPe=qxakq~CK00~+?z-Tb)
z7_K$8arM>w+wei9en2YFhC}h#NmoQDWS#F4qXlk9{O&agoQ`J7_UwW|Mzo{P8~iZ5
zr@L4H)Xi-H+%*{;s0e&6RJDTRx$|90Sb2c<^e=D2jL&*mrf3B-0|gJ@_QPOAjk63Y
zYko%awvvx_V_eI4}jm!W4i;T@RW(k6*6JvKNQBLCT;ieV{yZ>B|XD*hvnU5;Q5R;HT&!2YG?5
zFXQ2GE;Uk)D1g?w8^Xikl1D`4oJakB@`7_~nBInWRG`9a+SEcKDFun=NgXBN@@eVW
z@-3mrCa7)(ogCci2ey!1-&Cx7o6F>mwko(Y4KoQDJ?FV9kjlz&AH-A!W!h|3@z8-p6jI$j$OvpS31x@@bK$&b2gL#Wi0OrxmJ>AeZH#DS`+d
z<$Efc_YOqHGf!h&^l)LNjmdw!Q?^6fg#yO1`mlENcl1#dauXBRrBwu@qucz^gbk*W
zcE$hwCSfVFGEuRutHzSLm~e|g=S}qO|D0cxKluWvDI#nz7r7<}tVz1-?02zKmeR=s
zdEK0yM~i|O_QG>!(~{2FbOn*oqNEd_
z_mcwXw%g6j(wL>C(~p%ped^9YyZ?v$h>JOAa5s7@aF^EJ>Tb|>8QvT;+M4(R
z3S1>F26Y}%u1hMsXn7g-g52i*M;`72iSY~|*yU!$6emR{{wMBD^XlHB$KGw!8~N45
zWHLIOaDaNMVnI_cI$7@91{a0p*`DQ^i1bE>0-f91IrqBunBWmQ$9HanbIe#H&ah*0
z92n&Z2moH8ef77c_u+EHkPiIzA)f8Q`Z*_jpI#bl7>bsp;~^=Jm14Tl{+S$EDrEz7
zhpChyl9D(fVB;cbPT7a7PU&BflnaC{FY9_Zkk#bS(wzEd>wA;Jz9E=Pxi=7hUxiKt!`y){pRQniQk6wIg%i+PJ4i
z%p2kC%VHT^E{${uuw60<<8ZDPpE-ZYwa``Ij`mwnyg^|nKIV?b@Ys=^eCm8Oc#*VAMWBPQOs(672nYc%QVo3gPE80K
zVQISFYFLd~uSR(e&Zk8?w%#`!dIFhC&q?|Sj-c0(Bxc}(#WlJW<1QNJi&%4OhQWp|
z^&obp1VqJh83s)+tv-8QKfx%A-7PN#OXQPIe7cJAgG5!|^#s_RN9rBlZROqgET*bc
zomo~H@1Q9clwP`=lG^}fLj-NAg@7XWgZ3U0tmF0m`gGmK!cY1a5r5tG=R2LN`V_rw
z=@^hrfo)Vb%Nef?u=Ch|;EVN~n#zctR9i{4d`u^d!IIIINI7X#d|l>A8(aF4;&PrQ
z#Lq+#>tkU%@~A;F9R#X3Q2(9_UM~8fCk)ud&cTJEmrM+5BgFuCdyGRYQzWdXNccsg
z({NRyRfmGA?4mDly`6kSRsezcZHhLYqJ`~e%G4$(!zg=>599W!*RKLtySNr$2Tq(s
zf|07Nr7q>v({_`{N>E`1+*2%7^?ZnnIHzW56&^2pX`kA$z9Da(mrIyiEan!>L)|F98HFZO;>fC9}Lim&`;Kf&m!mvd*s`NYdqMyD>s;Z-VP*zf`J%!yu
z9>FR!nC%LNF0~wyN&NlRBS(e5H4?D>OTR^D2!jeRa{8X8$Y)^!*0XZ1wdO-cWbDe^
znZ^uD8~M*Yu5N$nvQLh;&`LxE=yaRY>5pYyc7Oj_X0P?v=cmBezZ)fi)Sbmnj|hm6r8MBhieDW9V4d
z311Gti~;N_>>RAdxA6*hX!j|+tA`HlN-sC^@Kwhx1d&$A0`h&vqBDN9jhHLnpJbQ9J!xf(BOPia!M9@CmS9?l7XK#LwDBCGaCMJAEM|544P#fu?Sx(!T8Z>Tlv+sn6>7fF=Qa
zbbP;$9y~U_T^oEi1X8{*y+(#V)aO-N&5tbRf*QVZ^CT}Q@x|tLdIo?@Dj4kG;LPII
zS!9;v+;A620uEhRvUkax9l?30~1n-8<0#
zS{x~*RQvJa!}I1CEQ4ZAYB68|?}GN)=RFTv>;pGJiuL$k=)Dga;M@rx>@9e{9b}f|KdE$ee
zqS%ZaO{l{LH}(ER`i(`inJ#+Ve~v##K6wdPpCxwQN)OQk1}6h-{|7BV(!W*pF0prK
zW!q_zeQY0|c|bUbEdYnKoxz+%Yx-`>e}?b47;DE%eJWm2Gz2ScHvuj@E$!Qv5VWpp
zWWdw;k+2Y$%vhj0{2W3BTO|e5(|ZF0F2N`gUi%ZM_Z1jWR}=JlnQHv{
z#&LY@WiydS7Qp;ROx@fQ1W@Weurz|{!?OZ#vh^p)u(084LOxy!45A^y9G@Nrf3DSg
z3wK$2_6x?dXHRDF=tSLnDnMV*Zw_BZj0=s{X2dg?+=
z{+#t47>4G{v**LndA0u1aL98nH0sRpKSixCNuVek%lh@Qnb7vjg85*!4u2Bc;l>oi
zE)D2yyFJ&zv`nSd?ckOLWc45%f8luwNj6PnjTSTQ@YCnJP$R{(q6eBY4@^na3eu?M
zF1lwd=F~El`%7Tm@b{Y*?g;GhG$UnLS}J3n$-v#zLX9Cf2VfL!Ud()amp5GJ6INUt
zi6)Igee5ElUMwUX=v8G`Ag&QRXmK8}d@-XAGm0;4>^Q=E<8^8i)Q=`if1LCg?kbsG
zx*jzZQ6iVkP&DbPB3zS30smkj%dOGbR}@OpLJ1kSSZff?n%WJOg<6!R`f40jx;OID
z6Ba2DlL5s`PdoDTMN>ePJo+*>Wnv~9@O8{mQ}EIGU_FiF*94Q7UC9_x(gRI!Z=_>j
zyRJ$;8Z*j>9U_HwHYqi(e~6WQq4*{&#i0wugW8faoJY9VN6K&zgJN3BD-v25XUlxO
z829r%_dWP2P~@jW@og3w1N(rhPRnwa7jKwte19&xY$`Rjsf0^CD?4IH@$1j4eg2vi
zPwY{542GONWLkv7jo(XsG)ep52}AL*wVwtvnBQf0?U5um2FrL&e?HgB2t|89JJa4H
z;qxD&pBB(yt8#pMaP~~)*6)_oIv5ZMGrc?yJ%u$FjLb0;BSrSa>fK4mNELo))uQiE
z!LI(6ZOR{(y}Km?N5wt*WG)2#82KnpyDxc%GVkJ2_xlj{V-pqm!4^pN;H%j)@KMKv
zvM3(w?AApKZf4EcfBIs%5l~UO`FjECJ(ftk`5F7z%L&C}&ux`1fk7K>Emrc=>9;0z
z>}mTd*W|Fiqx8m6{Bx`f_D?p0&EsitrC-#zQ**JiYs0oYXtw$@X|F?yt!aL^BrfH@
z+j7w)ihGXC4gV^Iy==E?ykVFKp}%3bM%>SXMlN0e|_(D4*fN7ajo2=GE@pGdX3b>nI0=S9!6*t%
zo%zSjIP7gxf5o8yf(Dm7qK>VIf|TinzS+t}u2ugRyrxRh{QCyGM0)v;06OR4>gYL*
z0aWbvi;8u4SEBtHa3;Z?iHY$f-Umw~8;t0$Pgi?A*pQ-m^{<~;QyB>MZ9UvC*sdP!
z3&wN^yQW~wPeS%MV^@4M1J!;+)%>uAC5NnWXY4rNjT$n3x#qEC-S>xt
z1z?G$q_)a|nAqp0>)a8+dTH8qt6tYWy3TfxMlxOcDD1H(%zqzX&=&
zMAdR@8TEEJ%&vTma4W$l=9Q4vw@|rLe$Zd5lXAM~dz*mnDt;J{dq=wmF_Ck1y1uDa
zv_rt@`6eIbnfg9-FtioY&lseh4`5(!VoyHA=?P
zL;6fbmChL-bfbZ+ij(s4GuS2Sj60pde;Xv>f)c(+8>m!zU)YTCV(f>txP@k214%Os
zs2RKOLr(QWJtFpDm6U^*$9Os6uT&q1gJ?|5rMRwyR4El5Jh19!R$fwf?`BlVI_&f)
z{h&|keViT#d&7(>mp}7@B2yjfDL2@EM%tC&UGoJFTzgmb+dyayc3^2w3o0DTe-wcX
zRmEdfULSAAClA#n>sEdEpSXfTZ`59#GwxvGyhrMB8@4TbTAa)gNk?v
zA7rbWoItlCLJCs=9m-W`iONUXUAh}oqj~@j(in|l&mG3C9zw?wUP5Lgf51ob?1Llm
zfzNyR{;Dnly~e+1nRad1-)Y$**jM>IAj=Xk64_ge&gu8E9(f7^NHS>;KZv<(L
znA8`uj=Jh7yepBpS+YMK{1ubD9C?ZFR5JylruEX);_1RGt
zWopNY&}@XCE-Fj?NOvv4p_pY$zG^Jdkc)&PQNc=UsCKLK<%qly#nqU$#X4>3{pCO$&#cTCfkw24dJzA$@K9QT*jIl0jV-Iw=sf#bZ9IL{_Z
zWW5JbQB~%7urB&g6^vmVw_Q`TUp^|Lb!>sc#ChpQzq|0X+{FmP%>
z8Io7=^7yi5=X@$Zu~6~Lz(=0yu3^4sGZj;#UN5tx@lkQ1REHB4P7-n*igJ^L(Y+X`
z^-J8JR9pe-e+RDjQJtkb%a<_`p{X?4nn=RmXgJuS`o{-sKAC%`hh}EiV9Y2Mg0@5K9>dVY
zK9W{XAbvkg`kNsLR_iHjj#d?s=vNLKt*YV$7TLM&=odYXUIw;c?og^A@$rxhN6B%l
zlsxSre=>JGz22_Eb%_S*!Jlq5dJqN|38c1ykAW4*Irv>4k!fl_L70`FM(&X5@-}>E
zU2W8TvT3n&zJd_@QESp=-$NEOe217O^gX9(-OnvmJ%%c%g{3WboBN0^NP2&aSA~s?
z;=8`TZ@ug0_Kf93Tb!$yIl&TOd-(I(CcGy^e<<&82Jd!llxZ&h#mgjzFR
zqXq?W=VZf+$fDWczs{dfu6!n(-ybV$#3*4sz>ZYwUk;Zu^at@)5U!&4`)UTpPdr+iV+-I(Tba|Lr&7L`+@~et{
z@$oh6YoCW9)tVGZm3Cd&4TOJ-t2I#T8OEV*pc|l#;=zdhGDq)#o@yrUA&;6CIUBE;
zPU4f16|NgAuwKSlSaTzlI1t%Ofb0Jle^WB^#Hx-?9hrRlX?ve~OMqUat*ZD&+aT5M
zMm4w`6Rpq0HlFI6m(WKtZcVENhxbVM#+Y~Cag-IjYJnv^KV_*sRmCwybV8upv4u^h
z`F}f;coQcY>B5t;w?=}|93^19uK)axIhnO^L7`k$3&g7iKwt-_NY%o{@h;K%e;6_0
zf8kA=29EHi;$3qd+ho-;a-2izpf1;#9#F9^C7t%~o)y5S)=NKquZ+lvSds1G^l#4NCrog0oK9
z1pyTwo3L>KEv!MwKoaXfp{1Z!@+n~z=T$GGEU&pv{KhT#v@h
zL?7(Lh-_IgI8h0^=*sD{<*A?r9U~w!nEo2Nw&Sek^TPWb(5kE8?ovmIfBsgs3dF@D
z;9|s7ew1b-_4!Vn>w>zryCA_
z#SLfN{x@b{Sq_tAASm7yuhwpA=`=BhvU=$?CIi2puk4@!E7(#)I><%etCQ;CM@2oL
zQX$cEPrdIFsR^yV3}Sfge-5kMfy0BpY90c`2^(?K3KBPo6@<6Jlp6N@`y7)#+<|{l
zNyYg?JO3-0mw|2()(cA4Jx91)iJKA+2yzm)1Xm9rlI$7VcRu
z>#s+X=ibXR!K}*o0rE~(l5^j3
zG|YA8@4iZNvkLv-DXWk{=*_>Qts`&!aL^kx{_~32CCegIMnFtuqG+KP?P9d(LVaT(
zHfKJCpHf+WUm*Pse;`1=n-AbB{f1c~!q=I6K{5&R8t%gSt$hKq{gd=gvG*F^Ig7Q!
zP9x%gExc6t>tP@Er?Ix|DT1o5y`(;woS425uk>VovGuwK6Z}nbq7ZQrnHcu~>CGm?
zPKC!2qZ0OONqC0N(Pkg5hWWh!%at~sCXJ2_g53BG1CG#lf1=#jhj8m-?}G7W$0;mn
zKZmhx0Oaf~2|34P#t4&!YYrTkM13G40%Jvj9>fo2_`6s>BOjvPT$VTqTkZFRSPkKA
zlRT))v##0r*G6l2d28$`iE0^*g4~gVr1GnCRb!xS7EM0(4Ebs?t*dp9)^u7XmRTs|
zQ+#wysF4Nve-zP#dP&b-&TKKii~i0075Qi`4yCb^qCwR)byQzx$)1s~(EeO?=1L?J
zGyafRnm45r^Pr4vbWf(8uH^WZ63mUqEK?V
zXz|Y){b(DNeMZI~lJXtCrmsm?Mucd6DFPuWPez@9e*&>VkWO1TkM`<|siVYZBhkyR
zQBa8%;^WwzjM}=RydBoY)yL65>{6DbIvEp~NzI0a$j_`w3U;m5NxVRp=jIxn$+TKA
zP4G4S7b`*jxFXXl>J<7J!C<_}GMz%6(PvJJ3DfAJ(~%6q~z_$V>zh)n>x5CbX%P9xjs
z^>}pO^?UCgH|FB+RzO0a@o5nYO}#`Fv9VY?V(sm^Eh;PuQ$ffaqF?KJ%Ct7zbVvkA
z2AuTTO)1cZJkQYuvL+^H$adwjpYl5NjXT3gXhrD8b>iB~Owx4?R&Z3zOyF4<_2PQj
ze{}pfx*y>&m)|qnv{NY(N8|YF0&@{}`O3NBy^vT2yu-zAcs4*<*}Kj&4eDo
zG)Dzc-=fvAU(zpo%u;Qe<9#ruvp?|DF2axyTntNk*gS@J)c1*n)uB4
z>8CFa)94i2&x2MqBXzIJk?3M
z9Iav;&xU5dFXvhM=%(lsUR8Fu`gd((8@=ShNMisAN8E`bY@9?n_u~8P*Bp$de`fr}
zwt4U8rhm&OKQQftpY9#
zi)_xAXQZ^=DT_(7pZXdlzJ@>~e=KGS5{t)?@(LF0Md2gY&Pl~-UlCL-Pme8CO~nXo
zZvDE24d`MJ!fLJFK?4l=!IhH~eWbpCpKa1;<-fJI+>Qeq$Z51&#kUxqwCa(I{lK#D
zD$#>^hACDKXW83@dlLF5Dc5twU;#-|E}Ki<-fg7e@4|N+Ys)B
zt_0m@vI@k9X$mhhE%jXFaFLHh;9aR&MCu-(2_$NR4p9()aB30!s8D|VJ+tVQ;BQs3
zsDVA(ZZl^r0-x(J`_qTPWvP)6F@8zC#_e|jzu6zRKlW4D9_F-4E;Y1pJ%wq>
zo$x$JDNjhHn_LmrmX
zye?7gwNT+uA8=pq_BcZw$PuE7gQ%!-5Gzhyzm%n-&P#_`8V!i31#HF^pxlG-8E9f8pF&*wR)%8%e1ZiBQ$Dc=zPjE#i|^x=*%WcC7UzogP!{bvzxj|i$UCa4>}aJ=
z0Wf^mBEceXG=8b=<-tphMbh2swz{j*L8BK`axd10)7X~V>4R~1aEh^EXg}Gjq>C$*
z<~TdTjv_rhe?Pwz9!1cd0gXvM(BXI0$SL!Y4_o7#dmE)w%J2sB$Z~8AY#(EYaejQ-
z=NnQ}Odzi6rf2(4tI6KOU%972
zMj8moD`d0v-lu7i8A1HEp~DDxZvG~})tp0qtRlhSfA{k{AcZcb{xspi?W>J|u9`xj6645@DmzdQ2s+NuEpVi`Szmd}6N;
z_#weTLZp~#Rzksg4ru%9$fee*dp~5SkBzO6rW4DMo|ywe`K2addpJaptiddF9l3f5ZuU`wP`ThJmM0$IyQ60}tJhqK}}0
zN8clyL~>HVvS#aKwGCQ?dxo4+835>>zj=jf>Za4oMwK|owt)M
zFYs;&^LiU+x#yM4-Yad|ho)_nR&~{2|E`exGh#;<%;}b{PeO2%Lj-L{DE*q?=C5WE
z&dhZKi#|A_b|nI&2ArJfN$?Mx-MlpCe?*%kRAn~ptky+DA~+3rR{5^a+MQlmwn?F%
zVh+De4J3J`zBeyu=++sGm=fBn+v@x>!`JFY>uj?5u!)Lgb6dHmkb;jnO$~-m8SW@X
ze-RPNV;tyO-UQu?D__
zGNQW4Zxj|?f@^;k*m!>VeXhW5!V}kAtDdYn-OP-UoUKpgWbUOH@#}e63{Xx*bkNbg
z*RM*Ue4y1o3
zjKY7EPr0l6ob%3pT-RvYCYgZ1{YqIc7$YNY;
zB#*ZSUbfZd$-{4a@urZ=f5JV#Q>Q3L$Ok%?+Ai;C;(mA?2}VC?v~vc5$L2i@TwXgh35C~L8Q&u}G(Wz!ESOGbk)c5uEQdjsPx_^7$
zKFU$3uY>qd%P(nQ!bsC+(=?;s6wj$8*UXcmU*1c~e@ieud#oV(J!mnn-l(z%Oyy`L
z=i(Ri%thBzf7f@QO!ZX!U5Kx?1!vRItrk*6V6vT#c#f0n=w0L{>}$%VKXLD5
zdbD^46Lbo?8k4bfbXIb{TLy
zKL^=n8|_@D)9=9X^$|Du)R|>=SW2f1JQjCDZ$RCo_kkj9Ho;
z1N?gD`0#5Uz&`N8BbqZpnJDc0xQw`4VO9_UW)vPuv+v8pTJV}(GaJ86=8Xl`u~Tk<
zn)__gY+XO2oZ`4=%!GNE<)Ks*eG#see{{L%Di7;Rqi24TT%6Gc2pOxwnlBPW`%+)i
z_F!fs_Q}q?e`838d@LopvYKC*SHULf!0ny2##%)B{rIK~iv{tGC8GN%-LAbHaxzzo
z&WF{OuSvSjP*#~~%2>ltuw7mcjnJTr_^lhYe=M_GH+U}7##`_a2MoqbPp?Ogi--o
zP2s3TL?pDo!pSa~Iqwz?*JsdP9@T{Oh=Sm?LPZlZX~ODZFYk{f;jzSK6x}anZA)qO
ze{-AgI?u0*A()KiyW>$V79!HdG)ZhRQ=cA^5Q0#UT4>reL7~7L1S<(MnJX+|89Jh<=m0L
z+M3-p#m1Yjf9Q5+aN-c@Q7$4Bk~#6ZoDLPGD3e72T_^<*I3L+ZLF;FiXR}s2e@N_M
zN6$*lV&~9Bn^#ev-Xgz*)OSCi|55-;bC6Ij1j(#}mRHEwk4~ta3!9Crwlf+LNe%XL
zSz2G`?Jg$<*fL26%S77Pu;K8e7TLb>2QW4_t(cUV20v7k$fW{TeWtNDe@P;`t}PI@
zpSshH#^{38iN0*%Dfi_5!eXC?PvELv$x9|EUwxqxbkJb(U`o06&Dj&YxlW%syVR=9avDI0kwLx-Jiw9);;jJ$2Z&|A)PRF
zu!q50`#P>>_VI?d$~Tn3n_Tr~x?S656oaMPMGJcvJL0QU_`-TFhvYbPMS#225YDAy}6Q
z3t2#Ipk!4+Ze(Kr2TxlxR6o7>r9ARwyP|`T6UE}xvn@P*G=Ca>l|vq_k*s%9EQD8}
z=gv_EsRaRL1XuiCD4`iZ5+vObd>q*VZvi}M=WnlL-N31Cf4>d^)~d)O{F=;Rd~TVj
zwPYtrDWa;~tkKu{#PL8hoLyoOT3m6^do%m+$r3ldr|2F1__XG6Pos##9}Tt-1c*Ha
z5awotV$NH;hH~1LT?U+SgA*o^Y1BBS34nI(*0=1sKMuR9Fl&*XZqg=*v+cKH@d
z8{fCccOyJ>e*}4zgWj=^j8ez(m~3#K;`Bh2vPW9gNAfB%b?Y%8z#chF>}rJwZ%3nle^>%SA(Hv9lo^Vdrd5^<%q(~#
zoJDnU35}v>snyRs2S4FNqNdjo4W3%
z_i|a5_$&`4pqFSjS;w}f!6hwv%tP}XWD0D>INM|Wdek2@Mw!%{D`2NGcxsFvY3wM2J07a=A+Iu&50*u6YN!IpU-~E
z+&5OE0)TGI8H@?m?i|^JyP(X)<2S#SgYSs3ik!C3XxCJhj5}n(oD*e|L&3L0qE{V4(#pe?-NpP-}wZ$3S^UqyI04rN-Z&eurfsy%;Y&WKI1AnA&5ulK6x$n54NE#Fe=O~L
znPT2&=!qxyCwo`!Eq63vz$%Bb1CKvliS+NZj6dVai2lUo+y?7>s6T+)>bcr$ay701
z!9h)k%|ve>2W!%s4^yB=R4#=YOY)=_Z>}w8s?yxN*?jLtd~z3fDM7}++VRzMVXBa2
zeKIX@eK;#OHnG#S@JJBqp{Bi3fB#xPD0<#%Y_Ynh5F?;r-l5vb$ZoX^dT(o{*c1x-xSYM<*Qk>9y_Lj&z3WZx>fH9FS-v}k$ScPJSWNxIK5-jTr0-spP9#NBiHfjRv3S|R)!b1hQzMo^<;
zx{69Y$X?I8*07oU{PzVC>-nPNmJ`~i#5=Tl)EvQf0pfc*g;;E*+Cf88!kEFzIy=!53o!$=CBY0M}
z%_-X^5a^kn&Unqsi#`H@w{4^fqSLd1)lmZ-MkMkxF`a&^e}uAFjglP$^#HyI!0dx|
zOIhhayh5u;Gsw`-2^%F=t;d(jH?|YZq;-AjG5}}bn+OX5RJRQ!Y
zM(igK!@!@pe*`7T_5O{o!wQ{pWn9JkRQwWMy8oX=(4PduJ9+sd{xF?F)6>$`VKpHN
z7#!%BziV6^CPBC}+Ei?v$YB8OkN2`XgruQ%4tu>SM_Qj=Jq5NKRg2xE1zGFa$VR%@
z*MCK%Tpo(#}1qe*}l^?8_~A8_p(_|Na*R*?s%4
zr0a9$Y%Z^yywxhRctPsLL=%G3X0qG-wl^hbSPOvu+}ANTQ6Dq2^0ubbS<02G#+FIe5TXnE5P80ze;T%MimR-z;vwakmzSYs)E
zvfM6bl%D0lxTWAbj(vaH;R>Gp)*)
zY@=bKYIMs6&z(UTw(eW8Kdr3%rrl0P!ELx!Bw|FegmvAw09px(h-QU?_BdW5_1|-M
zh~fMi<0y9$wtNQ(JPj^XY*mGU4Atire?O)E5Y6`Jo;;_4grF8M%Bm-6_Dv(cX=a)&
z9#}&VCrMhR1jId(U@(Q5%29GvT}ac}psINA6dlD5$U!Qyur*ct8vGB8?
zh5O@Exk!+;wwQKu`oOa|$bb118q`3GER83Mc24Ubf_Mz8m9JXwC=zOim$*!2e@^be
zG+egC)lp$9YS7~HJ;(BG$YnZ0TiSuKIv$Gd6l~%=cKb;FUokDE
zhOfd!u(S8Y99>R-3Xq=(S3Dl1Jld0;)l6VSU0_V;re=jVgt=K7h7WI8`^@Oco%m?F
zbO9@3Sx2gT7ZVfg^u9I>@v4`jf5HVq_zuq1N*;sBXa6(YKkP|ZC+b0{PlBO2P)ECo
zL&oVXBJCc^?3!5ptFJMhx*|O5(KBuF!~OkO6&=!ErQ6tAJF?g@>u>lu@!-}$0yP&aS-sxhA_w)j&^gP=PTh(P
z$32P<5C_8j79@D2w^_udvV_yUm*FZRaQ%Z_8*6#slcx$+Ym#gaAD3y1oyC|Pm{=UzAM_t2*C~;80L$p
zi6Zo3B=9}xg8&>rB+!z=LZV0cT0Yrpq@-fDnq6?=KnFWbd@^wMhqQHw>Wz2vZam5p
z|BC26J8p62T<3WD(0nsZ>5n>RZ9IFm9cwFykO7_(Vv-%`a2XVne=pxZQRO!J+wcJ)
ztPIBEyw0K90OjGb7kMs#i$`)1({U~Jk|V!^a`-wx$lZ;O4rM1eyAKxNj_c5gGi-rm
z)$!v4qtZAZFY?WSp*T)DHJrLRHyMgvP-(0ba=4kFPIEetg`yF(
zqilzN2>p{&nEKPgwV|i?w5B1}s>mJkCk9wTU-Ya_G%k{ck;EVPOauA;cr)_whhZy}9Kf1(egl)WzxD`gZiGrYbY
zcGhEPA}tI6<#l#vTD()yHQge!(jJydJ*V$~x^1+p@ywX0Oa3l#8ej~h7Q~YIX@RP+
z6PEKGR%D4Le;Ma1WAEdV_zg=^97?tkYEcgHi@ACy#oq(%0p(brHsNyiD3#c&cAMY%
z?!1QQ{`1^I$vf^%E&&k7$5&&|{#E=5l)#~Cd0@C7pFY
zrkSUmN;F_~B@bTow!Dy@I8;9h8wn$w-BFxEA9Bs!f5ID`CH_fe6$H)+GINOwI7Axb
z1tR^uX^|6}QVWhpk|fh***S!GkV!~1Aap-gZzCA+?^EUg9NdSo?5nthTJI{DI@NwV
z?~1pYbTp<$Xapal+y{^H&+{y0Hg1d%$LwZNN+Nzz76ZW_n?L)hBlzY=uvC32p6X#b
zMB)%qe`*4&a$i%*n1+G>kToh~1I2CB3-T-}-}g2%V~j*3
z)^8?q4>WCJ-!CIRTu7Ns5aPm~CZ1=2BKw+4F~Tm65dJ(*R5NG=ZC;JxaYxl7
zDD*b7l&srfaK5N!-Zp>r%*8TzHInjLn@tiqyp>CQJu$ckUE+c^&fgo<%~BJuF(6A4
zf4TR)9a2;h)yka~ky~>7Nh~g1IgmM-5*Ek<*&G;k{Xy}d5KtreFUh{Co3B!aKXpCs
zm9J@^dB5%=4@$&9!TQ}9Uy=8RNQ!+~)KI5crzB%CZTA&Y6#b}`f?O{&Ghwf(b7+xq
z$SIr3L^2y`_pFQbjwVBAD^`W!)U$Hue??0C?LzX-ucHJ=^_Ak6O&i4UyEh%|RSUUT
z!1S(hKkAe0dsXCjD!hTvkMm)L*-B~3;sBm!hvYb5>}`z!AeC&DQy7;G20I7*)8M7Sp?o>JTOE0_zE3_*J>0Ncp&e-?pC
zwVTSDX8o#X`&!Ncx(vGoigyAL#exL3Kb}}p2M-@%)m;8Q;hwlM+z~yF$8`NS_c7+r
zfj%fV+#e+e;x#d;Fv-+DcG(wFxx``i@fEtzDJme{HF*YmIH`ZPr`m)a5nzV)c=ixb
zmE)o3F`3c_5xXyXxmPjVLK#IvfAz8Z)9ZwiVUgaYS;HPbOz^5b2NCbWk6??33lnp~
zOd4qPcl7>DAB22^cl#Czr(oZ4)XFCE0ZSrqjjF(
z4P97qOpe&fuFpU9eGkxi=Yh5U3)`Y*-;`MQ%UTj8E@?ZAgl}Gt=jND$e|NXC`YDVg
zkih?ZQK27hjb3cAm$7X~&Ba3)2C#hAe>IR2;SLVrZI95n*G^>93nKPua+yuO%v#8c
zOzV)`Wsg@8mMXY1qD}w-Pl$%Zr3n*xektxTe((WCknq7doD#}GRzLrpVElJ=%u=a5
zpzL2caBNwvXPnN&wcr>me+2*i55b@>@qXo$qdOpGvcDg5R9Odoa%(Z6se;K@XPyZ;4GG>uxhN3e9bCC&v%FHICy<;=#SqR&yFI
zk-enB*_T47a%BzHG$2%O!2HNbz$+2z()?fUC3!tV!tGH08PGHFe`Oyb@u5>KNnEC9
zqrp;=NmF++p?{2B5{6$y+1h&d5wMna_kx3E0h!9Ws9w1P+K50Sttt2An7Kt?$Mu56
zZt+)WSP7@nk*|#bnGexIQVI7d{$Nvj$AXoqzpok#5}+fqKQ4Q6NT*$GSJK7QtzWjd
zOGoZf)_zR+c8r-padJe-
zZ8>8tkVnn7^LqDo+!A>`wsB(MK1cbK#LlqD6F8gcaOR^f|#
zyon8T-cFcZx(VT$of&lZse~8rlD6f%y;3R64EA4_&
z7YMH~dp%wFL1*n=io-ue9=za}7?^h5F_`p&9C3+-l={xbJ)Jo~r@W9;TK9*$NC*)`
z=JQNIV3_ftXBqaL7-bkZBd9Fc*3J8LVZL#_WJ6a$+9G_4OMn9Ze;SLG0gXK)S8Bc9
zh>xW=57HRo@zPiZ}I?tCGS^;J^MhIF6@i#o7?v)`L%WIMA$+2OqBSo4bQ!%#^z+R0RR2jM_p#kzU}
z{Bhh}y`YKvc~kTM4e@DKE{c`@}v~6xs(5#^|2Wk_{EfcVJ
z@VBOqCHCriTlZZ*3&aq{1_MjKcv}3I6N^7oh=f_7c*dnt*}3vZtaftIRvyX)HjpX_
zeDx3iVheu}(Du!g`VKZ%oiTT?>#)2wPMC>|n7zH{x^g6|)pXl<_E(#dmLW<~u$-f!
ze-J4$eQBPrP@`}+07Mo`^9@rzsxrI*cnH<-J_o(zPnxdCgJi>&45eIM
zQfI*T1;Yxsf4L1IK^B{4y)!1<;~qJe-z)hf`pWn)!}i$mqAa}Bpxgw$oLp)ToIMAM
z3>nmE`|8{NJ`Erc;RYl)%p6;?orp0cxuC+XOTQa`Qze|I@2
zBBIqp^~xXaTd*1sG40I9KsOm#LmS@TOR0%eyZ$&l`2t;6zE3hCRxI|_2KY-oCXBsU
zLEBDM^TJs^?}@3RK_W|GWj!vb;hVwl!SkiZ%E_Q#b?_nmw+0Q@nBa%?=sUDJs6=pnwe>SW$y31qIH}AU@CM+Bt$cJf3LVyw=MV
z>bBqFle?=9HOy#X(>@gKzP8U$8T7JJQQeQ4-RK@
zwb24k3+mXGZu&C@$lAWz4#Vk~#yD>n@!Bq}J5rOPlKX!!3mIWK`p5M+;k`1rJk$^QK7E9ck`<4yYUFm^i7Bt
zXr#+H#W~vi;3QWBG)3|Uf0nvKGxVsIyY-k{<&1Sc-%IOBFa=NW0<830>Qi^E*tOzY
zO513xH*>$Uw;#>Qyc^6xTHiC0_PnqXO}mSRjOad2g>PMQ38-H2m6`cI+4?lDIUQ@~
zju~ik5#&>+oK1*ZpUnw&30VeSyo@JgXL{LldU4KX%k<&kYQv;=e=b4-AWM4989vrK
z@$WjXYiy#-DDCCI2_xG>K(NpOTIGiNIcl&$eD|*M*R0$pUrXpo5X#c0%pTsrInQXW
zH_tXBA`hnm5H+Gkq4grlIX0-CyLJFs{tdp1+%{Q^)#g-;~(__opgoGS{
z4)~^Xu+o5^NcP9-tHuAETHyXn&lb=1;{g%O+Xs3iy44^(%{Q?P)PYs}#x&N@t4d
z=H|RsE7AU=JWCQVdn`np=-)AFE0s@`Xk3|j78E_PHvso!$jR93!t-Nc!-VNd+cAdV
z1oj(7gUs?Le}`w9Xt>A{Sf%xC_V`Pw(a8nB`n}wU&CSQAq~Z`JQV=(5%F>N?72nEG
zimT}aGL7~~L3dBc9!Ev{Ds503E9)sGQA@!f2z}+4f*o?v(rQ;-&`3&ep-Vl{)wr?*
z#*=zk;8ssjS9|$_{J1>@Z!jsrviI3Hb0AT)~6;F?MyNU`=0G
zM+}Y)igm!)_=tJF%R{MugMPiN{unhdh4RrrtlSOga>^q1C%zogFGfYa9&iDL6U{=9
zUoEBpfBCUji(=h)Vp~IQTlo6}ETAeLJwdum`c`I@v@ZibW_%gr9#;~&IO}taNeq1l
zG~y?7E)IM`SJ?5i3)df$y2~_n!=(TdsBOKmi0>BtSwO6uN;@~K6UkOS=vB+~n9gc^
zPWZe_(x*{Rj2Q!FG)*3Q^SAa3Rik;?X~Im5f9nPTycty6m+MiZdah6an?Pj0V)CHRad8blKzvD%tUq+i`dKvqc+x@l5rj%kPD=^?MUwxTq
zm5wPgj)T&jHIfzY%<_Yt`$=NowZq&5@1O&
zX0@$R;>WEKYL@%r1Ofowc@x{3v58w_76TH;cF6
z>IWB0E{6TlbbXGR-YR3IRR8B3X=|7U=7f3_kzkX
z)_;B@a`A^sJFm!x)AU(LFq(i@{ymM7tX6dq1ryfb`kqzDOa4b7vIaVh&w$C;QgXl3
zCS*~T4UX8U(P-(P``n4
z>&MM(#9#U?)4Nf;5>!vsgR`5`j1KsoJ%5U821e`VN-W21SYd0e
zXePR;q6)R@&sBm;-KxFVc;>uoUgSh^LHhmZaCGLL{6QlxHMwzHSy>tO4|f<}DC$1f
zsjnHodixO6DblO+yQogn%W+YUJA*HyA;F8O6q94Hv>mVODv&FC=#vGrep70DPJhzc
zmZb0`DlO2siMfIVqhXyk#^@1-%?M+}7mLQtpS>xGjg@WPO^#NPg}Dtbd{K#V>sF}%
z^kurgi#}QDB)D*_UxYL;{moQg3E(UBg+lHeUGgBVNC-n@(BZ~p6z3s(i(@15eJCj;
zJXpXTwRU*uWyVMcm7QVxeZ*h?F@IcM?4mFukV;c$XO>3Qki~$f
z@Xn{Z^ZF}Wa!t*kB>j_sw+_w2r=#dM?gnxstYz#))>xW#xzp`VrPNSfE;_t0)MwYL
zFP3Lxe=+io&bB(kojFtDt!vB343%Fs&Y0?r->ir3v)@g$uAat^TLdJX?SGJ)bQE-7
z{r4lMl#jLVxA1{l9DF<@E``WZpIF1!+2Qcjd__y5c#S)dN~V0m3#Qio^D-)@t{T3w
zrvZ*&sXt5sedMHi8ldnG`QMF<#ME|Qc*XKaM!q70`5XX_E1KQ`mt5+l)KBF%WclQ3GF!zBuwn1u
z7a9nLkO=y@juqtt5hWli_D;$i7HknM4_-l0HQ%!_5fv|GJ@Y|FS`{J6$!92B1Z@eP!H1OC;@6Hn16V0z?bD)OMox$*C_tQ#;)k6Qfh>n4Ov$UtHA@RwHruYJYXh(p#bysM?%V)JV$S
zsgPiuc0!XUb9d^LFsrvhuH|Iuzqe8(hVp1W@9IBvHUa
z@&}s6|H0Jk@Zi`%
zb@{KV;bA*d2~
z@0i&63_lzIDtxqVDzn#3f}s7_qf5Y+u{<0*1Q3x^LjmN
z!Zx#)cETDZUEwH~A+e=y&h~EYjAxM~X{U*u6>MEra7kC3v+duT?a
zt^}+gKP`1>)#Xs?JPmo3XAolV>L%25Lw{Q0zdzush^EX_vn9UhG$tJF44&I6x|h)b
z0#K}k4jrX2wKQy@$=0afYPx7d9~S*#4>9QL;R+CKJ~-Y!H)%hLE8+@f3WM2kFuMay
zxeiN=@iRjW2a0D2P0%$BO%anF)f9e=hkw?{Zh?1=H_EJsrUxU0mS0LUR8n&3N`Lgo
z=WU{%JrU5|!n9Zd%X$tr6VcbOPqK5f^9GE$ZP2d7ka(qxt_W=1utKY`VR2avrc_z^
zZ*vV){Sf1pyeCA4bR)(156j69k8l}ScfNDYE$NNZWZ)sc+`QQhmrGr5yP5hg>u;^k
zj?ligQ{|lh1BGvk5(;H*WOHbIgzQT#GQa9F4ka>goz7~AHd^c1yD9~fq&#eB3N1Zkf{LDKoHOg;xq@C
zdIA)IE+!hD4nQ^lt;t_N1?=p?WNP9JaRWh?)*v7qBt;zT;OS&-Y31^}1_u+IAd{{w0hn$Q&Z4ixu$i0>5_!
zpkQqV1UUnLCrN?-cG^Rvgd{=iF8?M55yItnO}oFt0nR|+f3&eOasDe;K}AIYU~gg#
zash%&KxU9Y7ZVp(XMpivHpn;7oaS$WK!CWblhf}WivMys{kzS-sEdIiE7Q00_BC<;
z?|&UL0l7MR{X?7oT(%h)^uSHKnr9RC9n%*B!6iC
z|2%2`)l16N&Q8h19!UGYH2v>D6MJhr&;JbluMlnEZ??2bU?+PMyZ`c8J4;!60L@jb
zUCgZhmiBMEtcwX`{zXBSc0kCs{AJSo{YcqCW*2h)SpPo008DJ0-2df+jHQ_^2+m%@@F;EUTs}E`QGOzsKgUI7yHh*xVXq31H{q2ADWGnRp_zLS})T
ziwoe*1{rK~pvPas0$^bVfn6Xe00&nWUw{SJ3HkSia&rM#M1Pz9M%(}vu|J3hz#{$!
z@d8*R{vbX8i{!r%4=aF0>JMTEu*m#D8~_&CKM2xK{ttrmQ}}}*{S^NoNPj=2KM2xK
z`QL~a(of|Ng7j1SgCPCX{~%5Pi^d-W>8|+)LAq=GL68nQ|3-WeE+&5vgp27P1c@~J
zH{yc$%)oY#jsBN~lk>O5-u_QM8!Lpc`9B~VfCcyma{Y$(<|fWofAsxra&$GZ`$rOF
znHGN_JA{RW^&c7>zoFYds(*4odRqL+=lY!vc6ItkK}d?_KOltm9|D{ZsjNI5tbm|@
zL_loT|9}wvZ2kculG*+PLUj2jdk9_oKadSV?oVYdNGb^OT>QfiSwHv>8AvktKirVT
zIQ(%#iaD4-t`0k(h07l++y7YqK8F5f=798a06P73nf*JI`!Aa{_tDYa{6_<-KR4e$vc<%}9^Ooxkojg}=Yt$}oDfAI6T##AAGv0KznS>!>Vh1v
z|H8j-asUwM0W?Eing^Q+gxI7tg_iqCew?g;qvB&euEKryUVkp()l%kU2NAwR{^nbt
za8Y1OKo)foSV4y0z%L~bRPvrGTmuVbC&g)UgAWNBF{c(E
z_&6?obVM{(g3=_9SNitTY$^FO?4`ZbvO+D}g_}KuP=DM7Qs^ZC@mU2&a-SX)dB?IC
zuU+rfpFZvD+UtxUKW`%5H7*$=MkbSjvW~i$VMacNKAo`0AxrY7q6Q2>i+xsJC%R>^
z#4v=#RKzE-;fsgWy6}oiq42-pTXOrFtKe;x4%wo&m*eMT
z7*xWMSD#TnbZ}TPtc1e7F*Lx(-V~5ak6%<lnC^SDd@aIZ()cO^ce=IG`hSVhp_mrB-cN^A=t_+WYxPlDOOd~3
z60eqJpYyVRC1d+=J$!H$i;ta9gdzaKn=4aez9?DNB~s!!Qg5&8EcBI&<`XTtK3CI*
zipY@s&?<3!>{li9+&L9~^N!kRQAAqaqJ?!>7fDjTpth@?@`VFf9i*_bfs9ud%P&pK
zxPOa-f@Cp!6f~j-t2nS~VAlmA&cH>fzbbQ;+@7B~_G9fA7$(`>5@P&%K>-L0eqRtfk!-V2&b#a{{ZIx+3LguD
zqL`w_s${Qf9vq?J3;|>54FE|+WJVu(K8}vSD)cjA64hzb`4-)URtMMDlk*oUXe*Ue
z{ZjWf#eOk`<@!pI&d?URf5AE)=03hPKb4nCq4K34dj!
ze?)@eaDH)>#S=G|!0!8*gD6wSz2QXsN*`Rdxb8>&wNojm7b~qDH7pM?5PVj>SRJ$(iR7IUnyVO(tjO;v6LktK-@~Fyr@8}mn4U%UA-NTQ
zQ8*$43H1+-ZtLTTG^k@9Ue?bxH=@AxJU)uefY
z`#wiGbiHKc`5@|Il7Behg6ah4@+r=JvKqB*f@Fv&zL(G*h}Gv!`s1nk-8)N8hF9!E
zslRre1$c-qjXAmWHuv_Pv&cEOPP=+-i|delo#Ne{HKs&(u)pT;cWp*Jkgf+h)%P8k
zQ~Zl#Vmt~yTkHE|y2sRiXV5UZYE6n^K7h%Z2Y*tJ6r#XVP#e|@52lto
z3wGv(6+V8WG$rfJ7B&?X3+!kVEV2nSnd9!8geGKMmS0#PRK?vu(VK8(buXe%nMGcW}zg~FKg7hi0Q<2LaG14n3>k~UV~BcOk1YD
zpn&7p_hppI5P$B-WEYAn*$%`CV-6mJc8;uStyb%poFC8G8=5Q~V6r295h-1z+3tBO
z@(#Z`JD=gbdn1qZ3>)?AHqDb)+q14C=`H;i8l&%r-12BK<0U&wdang?*>Ea`nO=8_
z9eHD*WN-*UiE2hfD$RZ|-na-BS>vg2T`@V5rK>azV}GhGT<}cCb-G@l2_cw#m)J;Z
zIxf;@;DK(u5_C+d{>?sr>+~(@h9Jhc6h>GboVMM37>%P+$8!C;^`j;Or7ZbsVPzqf
z5w{qR7TT~5l|aE#9tVYoRi1La%hmv=M(rn@vSj(U84E_5>rzT$^^5uTPYz=%N#L^P
zb3H37>VL))MLlE8+2;tNgqR*qerQBEnSi%ik679`5$Mi(`BD)YYPEX{LTS350&1e}
zkqI)xHZtGRE)^0-X7^#)eEV2th8D@lo<_gpA=_CayA}W+b6zL@Fm>5$}blDf^0q
z;2hZ4g^@1nq;?n05gr98)<;pEIU_g3aDo8(p5q@<@gF++OOh2ccWu89aGWA{1BqsI
zk(_OZhx2LH&*;7`@GxJI%iQJBzVTA1oq-wLTFW
z!C#_no_(mZedb@M%GY2)^|vBAd*Fk=z<_*`j{^neECH(dU?XXVBjV4
zh8@)vhOJTUG5-8*T^}#r>l3%A;-=2|DOI#b+fzz{AtH*K_+*2s;lz`8lZXDYjTU!1
z>*oxBU#ktX-P^RIP_3QZpv9KOrT+cCsgPZu-LpMCq1Cr&FV8}lcHLKRoKhE9?|+|J
z1d?!niMktHU4(xlQuQ7SqV3MD)kY-KUb$aT@?pg@$2UtA{8(v#Jpu=XVVgTQ1wxU0
z8nJX83{|fPT#F#Nif^rMs!?%dLKNWyO2rhIw=f9`ciR(tr#4fw=f2`zMyGU&Iydqd
z+^-xysxu1X5>x($u)2$?+EK(vBY(gwz6gVfZ5BRGze403-vagZhvY}qJV)-bAV-dj
zK{j6H0_sd`BOpI@2xB_CJL8+9pnXfumu%_;@AIjLo5cHxqUd5WyoO_L7+0dQ?-83M
zu}^rq1Iw|n!k264tNzNChCk6hH3MUA!iGv+mktG;EakKa_?}p%%zpW@BY)Z1_$eWh
zn0)sVUg8nX^p{`$HK*6vbxffNZ&drdZx>a{JbN&1A^AqB_cIDxgrz%px)>>VWZedt
zd?kaaGmsET(z^%MhuUhxg5ymC>(n4B6W?8I*sP{VN+DI`D8DOO^2A)0XtxZ#-QYZ@LkSxu0&}TS!@L@^9mEhl1r|
zva@*Xv(&SGK5lnsT>8jymgH#Bptp~4AmBJwVQ0sW`^O;gM@!FoJb&Yb*Sf_WJoCG7
z?g-QNi29*Sck6s{5^iJZEvPMLub0t%6UP0^kRMd^Q%;dbk4bE0su|VG5Vql0`jPsj
z^C@K+^XkuWm7GM|xveiZyL~6u3{uLw>T1^@kGZJx^+cA=nS~B380zzxP*M`lRN6qB
zFgtKt#T^YAEY63GcYj}qM2Xe;^#?W^wLWvpXz`<(U{MT(_2m#{g%9X-FO3S^epPK{C4CYm?UbW%$AI!`pxut6hJPgS9yzaSH
zyn|7oy&#hQ}MOhpla8)f6J<3hXD
z1u3Q(Y=0aohlrHEY-c6MW(V&{SlH>ehr6OF)VpZs*v>Oityt74B^RxQuf#1{yh4PP
zXMijR48=L;`_PM$fcHFeEna@YNes4{>lIy0<3_{X>Z&Bv&*=joyf4`{Ut5j_`y8>F
z+&=pQ!XMJOh9r9A6SBOUN>FFN|B{d{v;?PxZmrk_pvC`f0>8L)ks5^#
zfqx`}W}_Zw=&WPG;t&U1!ra33mQSeEWU#Av$Jgn+@R|eFK1rsx6{)`L{g+8o+1>p1`RUtCVK(z@w^*Y?q=j?YTlLKMkm@bH8MT#IR)AdhAx2sJ-7c8=BU9
zJkIyteMr2gd$BGXnbOj~Y_WNz*@{I%;DHyMIsq
zX1Ae@YD(CEK=08dO_e4cx5Iz>XJUH^&}y$?U}inZePPVW%95#5=tm;1tf4=3de#kl
zC`Qdf4Dw)!8mK@dS!n}UY!@evgXq|y7AksmZi@YZuRQ6QocRYEd?<^?$0?c4a4~J^
zqeb}}?qoTe;zk;zvd6;hAFJ~m#Z{yT{rbJHny4N<+>zsD|pmWI5)UDu&qKU(UY$OlVQ2yh~koqM0m
z`%1~EY)#nmV<#BIvNb7)#S+G$-3P{_DW*Di5X&7a2t~j#KS4N2>0O;rH
z%qEh66>#SM6^hDg(?y1VXnzh(i>a8}MZ0l74rh3NGskP@e*z4UTe|X}o6Or$Jr_&S
zi_#}B-g(4)n#NNiOSE2#3@Ohe7T^!-d6b$neAX&;FmSc)F3%2
zKP%K8;n`rCnKcPkfJv{z=fxGmH=rC!Qvg;uUKnzNw~nj+fLLw#_kS7d4xsGfBvtB<
zWI=c@wT(^LSQ-@_CpVr_zPze+`W#`*-0WQSlT*Z6@(Au(iO$i_Ogaouhq#oCj04>x
z>P2#F`cE>0&J3u3C-WQalJ!A{MsD_IJrD@-3PMj)$2;h3tVO%zsw
zy>ZJJ_5;TVB8`q3TZL0BcNdk^v5`nVg9cQvk@9KEy|6(1ZhmbrKjWA7a#pwD4%ZwZ
z2l@|e#4bl4b$|QV9ox0Zshtg--ZdP;WJdyb%1(7E%`+k;Tc{+`Y{_)71PP;+=``kS
z<0xBm=FXdV@t8WWeK#sKp9?FLKSj^k*Wd#i?8w)5rbl=*h69PIEnCCrPCo^d8T>?K
zv1>YPWfvfy3y3kZS@lR4FDD}(aA&5*GjrM~AbvmRlYi1xW^ZTALNhz%>@bZll#+b!
z22+yt;m+pRRDGFGSM$dbruXnKo}n5P2j}>{k)A%s*3Ylbwcy^B;yY2X%@3I0jk-Yz
z@}VhM*avDoe7t5|!PK}k6zA9)Drg_Zs=v46CJW{N^l8-J_SdS|*08St_nExx`JzLt
z9+|uQY=4%#jAo$E+e8FK4T>P;#K>1s12gks$&@q7jQh#+__-zUgl32Xs1GE0D+2MS
zH4=v}vaWur0I6U-mz=!b2e5Aa0%iGwpobXj=Quo$;9?h3v?Ba908eq%0*WgyDJ1UIDaS|uhe{}(-9ouqxksIhvbW7I_vu~
z&6knWjCz(yVsFzK*F;M${OX`GV$?oC)vxID(-xh_
z$WlZm8^W@T$*($d1u0x@Agam;*Eg94y?OPa@6>*Hpjw<%_gvG1n4sI}TlL3I_q?`IP=yGY5uM_9CZwvPVP?$tW4Rr@+Q!{
zoK&{KKh_P39g4wF!5Eg31%x(rTz~UfU%thto|SGSELTtEWFQFyO2|j$%$2gtLpSu1
zlRvSaWyyr3<(OaxE!l2hDfum^hjJbGz@Z$@iKP1RO-xCFK_PwXk$1jaPyrKBLrkdD
zr9YOz>v_KYd3j8`HWv6AmjLs;pjljG1R5v|99)?G5^Wlh71V%DGVBL&B7bj8KM${`
zEO1f`UNEMXG*6+tBWtPA_i&#i|~2{R$FO43eM>CJn1nxu*0l+KMjS8
zLiZ70SY~loWAEERRWwyzqC}MxWgTx$Fzxrt;LBq$DLe;%P<+$9W^J!JCmtWyHSi^O
zIlOe1@W_-P^00e$DdIhLJb%Q(fLoDYO^b}$UIghh;w7D~*B-_Z>o#Qe^g%uHmTZ{<
zObCsY{W3)I=F#g1VQ7|`IG*&?ZK#Q|2a1iIUu=+h3sbMmvCcfEqk$bHWsZ@}N^URa
zvMvD_;Zc>+R|xyTu291^L_>nJ@H!{(*ZVVN$jq4Z)-b~1h$+#g*1u;`9s%*kC
zR-HP$OB6s@pT#~SRDa$wKx7A`pGj9;%7~CD)f9V2Xtw!&4Km%KXB~~_pE>K%<(5P3
zj=$MroDL7W-Q-ypNdXlRZrrTp@gHQG1&pq_wFU1TTLy9-WSh%n->293!vBCm>G~Go
z>iRbJY=!jw_-5-ce3VZt6A-gb2hj`N>)?aeZpv9SRx!w#;eRk|tZ@9|8gT|CBjD9~
z%ap#ZkmJTdFmi9vXswZ~Urgl|>;*AK6AD+i@5sdY{f9QI7@ND8^V$hAbWeTXd@kd4
zvxLqAeWiY-pK5Y!TA!ZsF44?Ke3m8bisTAwC3uL-T8ku6OF^s9mJQIV=K;H`NNc-O
zQFj*cLW0&xt$&0SDuqXM_?=&;wy3_GJ7-C;5Tt*4sEMp4e`EDAa4x@1(HZ?V8*3hB
zw#AcSEsI=uFV_ej0#P{A~CM4x?RZ
zS;l`1BTm$!M8`e2(UQN0$7zth=VyQzn*5lDzkeNpXI;UK+)V+2B(K=c
z+njbtt4mmYo#HPf2e-WTgRYx)x-v`zEwoHdt>nq#!f+e2>6s&7GC$OJX(^sol
zc}4?VHo{A)y--faOLtZDL2|*d)qD(WQfY9ZE!j=u24`wTl1MfPZ;?
zcSW*}^ZeOl*H}%XQa4#=>TAo`H|wQD>RiwlGa@3_+kLbUfkNYt(pJnc4d1`-M%8CIfav#KShnbe$mINQ}9?Ro|W%
z2rTgZmY#9M5e>D}_J4ecI9f~c?Mf|9RXQ8?a6#HJ@Ro8~Y&(a|qf?blOB--Qc3lqWTGYNGiNUjxlT40?JgJ`M%dA{`^I8m0z%+_#{5=>hKD(@7x4uP1^$N2{LB{nq
z+zH*QeQBR@jGeAA*f_CNLL9r-v0+GEOE5r{n~MJjO=R;YJ>(g}njUUf`_;WuxYW5$
z?P+el`bCykl?>*`tUfWlxPMll>q;eVql_Z+ve8dnK|VKo?mZ~m)#Qc;7W38$$zy-SOiBy2r9I}(h+6VN!EbSY(Rwna_QNT^(^^wxU{M7F#$
zqFh746SpUIWrT5`v!?#sv~0E94cWj=8HZP(f?F7S~Yq-AdwNd5yvSU)OPsL#b(n
z7>`EcZN3j)wh|(Z>bbN)+Ls8~Jo-{wSUu`_uJZQC*`BBEcxcS
zuFsU}f;q7D@tc>F$M0Desc{w`+yb{|;Z&|pp0wRK(u8!8@MIa{eo41CwgxNp>c8>_
z5QKec?le1^3$B2xP!AhjkN=W%YM!c66}L^ptWUTx$UaYvW}E0qGC8v%MT2JYt`!c
z*Te6cCK-W!EHe}!v9Hr`M`^|)p9)oW4g8&Iq&43
zsPxKphe#VLY?+bynp?X9%~8I+GJAo<{C)KXpH?}bS;p1|t1jSAZA^^2o;X3f~z6tx-3{uaLX`hC!vDVLn%&4>`Od)RvrAT
z8};)Y4>IqDSj8M?)))78Dfo;I1v|l}epFj@A%Cb=XVq82H5Q<97S1RtLd!^wjI*sglmg>}-f#eB~T`^1`f4-;p+I(+}Xe|2@VUYWb8MHy%
zvsdH#IATiqNvAQRvu2atsW$&I??J2sBh6EIayRg{KX?54HHDqAv@ENz#v((0OXh5q
zl7G1Lo@7Fv**@%iU(D=F3Bev#y5!M%S6c_M(6YBimwli}KCNn-(TJF!|4L-H1VyEQ
zeOK%5{M8p6$+%3v9>2y5K~q52=<&O)LF7r0i2dr~*Aeky-lhkR9oO|Ybc_L{a(77t
zH+5VHw(<$6twHw2yEBF4Vr|G5PC5Q
zj$W^RiRdU;IZUnv^T4=pZZZ#Xwtt4-`4~f&rfhXWU@8&kmA5G9kKs!Not8k7uz#6j
zW8|&H?&3U71#$0+P!0YuU3?X>mZQM2fAD^8#
zq~=h!C4y|c&XNo6M|)zRsce0Il7D`8afZD@R}$i@+9$QbmB=j9DR)(<&Nwo0$uG;7%V1b)b(RCbnf$GbH7lILozw&O>I?~2P
zyicEw=yo-a`Lig1w(p^fIj6tTUzA?I$EZ{%Lrye%j+L*V`;b^k
ztC=#RBLKpBo=6n9P9!QV!*&y=D!dzKM4}I)TLOpJiH@`#?^^&_NFDt#LiG5NvHbOA
zx(Kz+yWy*LLb)c98kd$O%75+H*hBA1GQa!jcV8!yxe3)%gk0IE+2+sOlj|4Nsc=ji
z7J}c(qZHM1jfNq87UenGmM)1S%#FK8)qUt-6>XPj}|Ccz?-7ZOb^b^QJ~FwmGN?F&)r!3PhSq2rkJ
zCyj@cQ|<<>x`z)`ynkz@y>kq5QG|useLq^QasfzGi$;Z?)#@}Au8t2wB)
zj2Azdrdd%r_ozN+FVUoiIsP0vO`#9mnC4S*g>fGdjTMQ@X
z3g2#!(SZH2xU>K@tLd2X$y;%q+M-idF#b6pP9h
zTZTxdmX`d(w~?1u+ysd-ivP!d^)~IeSByBETCRMg6wkA*t3G$YS!`ZtQr%GUj`pSq
zc^)BVTK_%>|IARMeWr&f9mDZxf^4j
zOYiI)+1vmJ-ghb{Zx~Z#nnCSLn{pFZG|UO5IQd<
z0oZP}n@WrOnuEv4${3{>`)F^h0nOEvgw3qUtfS$}YJaNKAuw_S?`<=gPqe;Cteuo9
zN+6X+phWlKZal4x#2YShY!)6H<}I6(bt=LkeqksHWu*{DcJftOYndePe<0%_8L;Yv
zp@@KUB!$*5z|=Dy8#6)(T1npqKumEeB*!amb
zD!@Ks&p9DJp*&*vjcSoA&m)&@fNrP%(#rCWDuFeDH!&lbA4;^&r0dOPmhxod(dNE}
zC50Xf&`$3>u(s#r+
z3FsOTvapz*vva_h?14p~;f+?h4sEALX7>
zN_@o1NLiu64#!9|6;2~JdG1H!UKrEyilj!wNa*)=3&!+CgBH4is*Q6sK{PSz!#rsq
z#D5SC#riTtj>%vWbRwg36+%*5TQ2GkIF4Er3zBx{8hQ~5Hj|&VBcp?1Odn(pM2_M4
zx|+|)n6(ad4*Dc>54B=qaTN%*c#%h0c8dHNSoq{8GRXa7Gi)ku@Hs@V)um|TYaQ$K
zLO$)6eQXu~Q0XO2)Pjf2ld)4v@JTbZEPs4K9c72lyFkdSaLria%E4b)jL%)^>vSmYSd#NIU{Gl~VfMLBg#xFmvz_SIxM+g?E#?K-{+JExa1^)V`
z)!F3~yC2Af)Vr%14X?$rG)1&qL+P=gO;4hPm10V9U+pHb7rDyYm%tzSFr5!sf8Jg;
z@$ae+{lH5y)drPKb@ojw+A#zq`c6gu@Z*(^Y&dSU3FA=_^@-y0eIfd1Dvk*`3>4%l
zg7=jdso%=+GxKn5)v{ZYv;Q(4K?huz
zoye)Xt-2kBs^%S!{~~C7wf9azYmv$-ClI_sD2;s{kYUVdzJ0Gr+P?p3!*0~@2~R}(
zTj1KX!-+Ub!)iL4_Pm~QF*Wl6*f5qky0vW)+(y)iECWXvS@h#wcYpuQ=CMbU0A9aH
z?y*AJHAnGslsu(2PI+pPV$4Cm>@w^pyB~Ouz#q`dj~us=h5hv>V+!hcu5bK7&*3i>
zfc4c1?j-$6TNI9NOC`Ckl{V)RIWgSeY2A5o8;82yFZ`lOQ4$z<#H$%kp=E;T0-NbP
zLiqzXtLbNY;?r!Yjeps4m|we2D$jrAZ*AH)p@!QZm5XS3d*C$qA@$YM@~#E>joZ__
z&(>3CxW6a%>vB3QEI(wIq(DZ*L}UG2y7z0+mpqD?hwIT-)rV8=c_5Apwm{B;BP%#V
zdSlZYHc&ax0a?2Ln-d`HM3W-m^T2!f)SMp@CYWr11dR{kmb%L$9An>Z9r8)&B9Po^pVk*t7y7Z*EJLYQ#Rj^CmL{#MH^E$$F<@$7KZqetbWxr=4#AP9v$$nhrqa>szgf!M!3Rf<7d+bAu&ORU%>B*9Y-4H01uu);X(dlaK4E$%@V&zMZ
zWw`fSH)aWqywyuM2)HBXUri<6J{mAYya`?TsI>fzNlVixefTFhvk&dsdROZ_*GaiQ
zJ1ByRIe(O@g-5nT&i~iFF+y#L?$v!fHX0^~c)l?_s^RDQQ9}HhFv8|M1>)CXwsBqr
z74W5Y@+8eTC%r{cpAUM2ilh++O`+^bwPrxO5G4@Td|%#Sze^`9T=7+tC%%X0iSV0q
zi}OuxKdEb@XeyakLs(!ND6jc6U=F%`miKxWzJJ*0J3|Y5p_5Orvv*cb8Jb%^BgV?|
zfO-QI%~#CFGjsCIR!nP(_O0wTdI;BfCYxnE{i1NcQ09NW3Nf}<7add?mlfJf{(K^C
zf#xvHM3$!h^Gy=a(T3q(2K
z5;yC}y&5b>0Y;iMe`u+B&S`@-aP;EY6A#sscBIVeNraVRN%;)ZGI7hX<$Z3==6_aU
zvavRvvNp_AUD|!<Fv(j?n`0}bDMvd
z7pJkIxUa++#O#akY%tQMp$X13;PT7HRd&~;w}gtQ+X#)i&xCVr4_
z`fVEL2B-BczeDGxR|wEv4WyM-s!RP!0~Tx1X+7c}U}3v2;5rz(AXHc>NGB>Z-)rXk
zf*474t$OW&pUM7V3*T4I&365A;YNQvu5{bH`^Hf`R}!9;h-w743w64`T(m=kUoF*WB7Gb7hff;do6cAK7%Pr@P%Q}oQ$=pLZ
zE2LkVdnzRBQg8hUF>LjW-61L$seRty1v)A%o*XGj17{8m+jW&J-%tnl1+{-HY%7QC
zS6db~_X0|v=@jrFx0kVVTHt3rZ4Qd}YZ@#8&Aux@jwrO!Sl>ZK1C&XLo06u_aoCDB
zy;#FldR}}oGY&uXfQ$L=^tdppAH~-k#=Z92WdxxrCd2COGUc(1Bcu(sc|L<~QR^A5
z9Ym$LSf)xws#%lGmD+L(jKhD!W(Vl*>0cPCJBhOtnDLwlFoJ1-<`83{vRfLC=F0|U
znY|AL8#ra1LqZ}*UthE8VioNnt6(R2fX7CW^^b(VFdiU%Hh4X-i{-w_zcr(q2)`1i
zqchgFt+X0tVZ<(7mBd<%ONe$CrR@Y0(7{B=%4++`yp1d4!1RnvZOwoF>6IDN99rqd
zkrcA;wpB8Fk@&QWQ_1c&m_4(sxgB2CTcl|minQ^>E6)u4%C1i)trE5Q?HvIntC^n5
za5%R5_Be(>b^d|N&&@bvGH|5XDKA*pO
zwNq3Mr+6kXq)qmNkl?jQae5Mqu2>qM+bIQ$+t6f>CwItv+<_`x|69sOHvW^qHw&*C
zPrqwLEn$(V(PVC!v8EFM+!w}?X%|VxNMn+(jS4OZXJwwMj_H4k{i9_MLc;A$niv&P
zWum(=3WU;dB{t)kFR+V~A7eqf5u!7U5|xVZSy`Kjukhi$Lf}0K>84{pq9)JbN-ND6
zO>uIT%XVCioNm}E*s0gFwN~Q?dUC>cJ6ETryRR;bvY2V};@^0Dl3r4BH+_53M@y<$
ze49=Tk7h{S70G`xb2mA^FIJkS;h%aBu;!AEQRQgeGf|>lENBp!LfO6)MS^w5Oljlp17^zvs`~!}6;Dq}qb4nnPK?S?J+TG6(b|
z_Ui^qD#;ySS&g5*?;n$8-d>bCvF4?%4{vEv$;$NcNsfO7fiUC{U+YoPx;x&|KCSMz
zBf5nN%2waKp~Wx3H1OHe-3%|ogK3--@-HO=Yy#N4)0Cq`qH}hXciTzThM(D2IN&-o
zbw>zoeM&DK*cGL35n|<-cfP?ydMTSxP?O5Na@+Sp1x+Z@#3ACY->L7ycq93>bWBcrk
z4>8u+WrmYIJG4=r$k8#r@l+_(dbuop(YeZX_F8}I@f*#Gu}12Gors@$Xb5N@i{IqE
zPy{q#{(2<{o?Ggzpw}+?+^X7jjQufEUmlHqCF3H0tY6{94f}h(Wbh@
zq8)!h{&6mMp;l%&miqFB=20Qqx|D~lYi1RJ7w$O5dPNiBXe))g0;oXE3=+LGhU6kI{tR(7;uh@oqtJ{V3i_%imm;+ORXZ=Qo_J4jU?R*qSj;a=M*+9v?zZ;
z5^)isMdvJQ_D(wgn)bjvJLqQ@v#Z_tZtPtZIDGRYA
zvy85&$dPu{U!2nd4lDbrakgf6GBSUe?cg29v<0;hQWsM2L*>-)Fu+E%klNtFavV6CEf
zo`FY^!6rp1Z;QQ
zkp<>huXb6~DxkD`_Q?)8TG?NP=|UZ=dl@;V%^5!KYM2xzx`s@_lOz}aNJkjecjOz>M?gb+)$m=N;yo&M#sI3v+*H1(kiiP
z&AS!lu8e^XW~pZUp)!dld1rq#bnO)$WhzVMQvfVmJU~l-02K01$?)%#LML)r@N;5j(PT;OL0yw&drEiQY}*$_fr>-iF?P-y4I#3GD&|vY+Nbp&jk3l
zZiI#bW|7A^G8yP~?NO}7;1^k23@h!nlu#ep$1M`OfhOU8>FfQfpOQ#DMQ0vf;{5xi
zjJ@t*U$FLQgsN8v>T*Di9a`{0kNY3NPmSF4y}6vY2%4@%<;%URWN7WvFT5L($KDLf
zwhYjWNFmWDJN)(;gW`V-+MZaa5=CvPMbNts)9hWrar~M@XOZ6XdVaO0Gn54l=XQ&m
zgv0328O-WWxPjLW-v(%Fpsso1eYn~-`@PzzaoEBIZW&E+EBCK3NelF
zNmu|+0Ng`it~gns>`6O?o86jf7p;b=K}V!^;~JJBTpSzaVzYXj4o`QK=QeP?B^Qc(
zdxkZ{?(nmKn&O%^pxhKz$u@UZ?PMi9RCJievGq!hN~0E2);?5fuoD1$B?q%OOcRji
zBK#h1P%36PB(#5L>3|NKOFdd8C|Vg2rR2rU-3mn^g`zGS7h@!=r%x7Mf+GU=^mQ-(
z&sG9k*bg}R>oo@_=Bg%-Aqa3V94E&vi-G$~FZw8SSZCBmfoUF>{^b9UjdNHM1wfQ+
z*|u%lwr$(CZQHhO+qQknw(IpQdNa{KkrA13@?^c&YUF?LyDQi_s4$QguVN(((LJyK
zAj4p=Lsl=?mL@-y?DJpe;e;0TG(F!sf??Nx1au1m5(L{UtI#2mn+sE@G)TGs4x+Dm
zpuj3}_GMjg*c$~2K)kBO
zpPQfc?a=mqbF#oy1!$pY9zDxtx?6LUEz?%7le$7%gr63&->H
zU+P0C9|(79`@!mUZw+7m#pM!N@y{{(2?Hl_&qs%mx||#F%mz6O1<>n|LHS}sd-17m
zwW)urT;JUQV2|nr3@5J^V{uaaOP?CoQU_tIEXd4K5n}1kc4x7(>1M)8*RHYoKzg(m
zHR@QTlWJzL!AjBxyR-O$u-|Q?SsC;~FA=ba3D*7jTMI#=sr)w`?QFs?rca*CSbya=
z!AHYLMlr<77ttQ}aV(vzs&|8zo`0OL{*8ZewjO5h`&m5(`lDHAY-d(n^-o9@aX#1Pg{v8H&S9sya?w3DVA5NQp!h
zGimZR1h~~Qqxr!VQWTD;T=th|x}}0ERZu=DQdh_*R|BkfCv|Y3_$!*bh0wbxmIQw>OZ5VrbcXuNZS?cdNY^MlPg=_#fjue@Zd>kX
zLZsl3iI>_QY9eo4D>gErZ*XOfXH}Br7;}EW>({5Ff=ZAc8J$4Q>uZLJ8`XtNoxwj2tOYs(KdEB
zOE-o42g2fO>Ra{#{BM>`pJ$U;9rztLVF|)TO6gy+I*i@KN%jifrYECJpAy0Hemuj7
zAvUv{FJ-VxQHl!ao27egiDNPM`vsW|0%BM40ToNX-DI0m=grDt6cvA;BlPE_Jsxty
zlM+RN2Fe?_1aKLatS4ys+A;ab^8lM>mJD(hHsJh^6$qbG0e5I+A(^dgRO`Ia96Hzi
zS^~Hl6FTG%SZ_S{vR@h8!zpy6rv1e8Z_{WHkSMULg!oZxz*SpVBmaW>O>imBgtF#29MOSj=)#zslyUvUcXYp6^-*eO?XTl^ILXbTC#Pwt!
z*G7~RRK~Zg>yz24DmQ+m#+i6Xq=j<9~&v?Jt+oDztCJIJDumYFvP&`IXAKA+4
z@b0`K38=Xlvo4^!2AQncigIrspRX2^l>b0{gb*nTa_6r;&T)TYnuR(5)HROs!KBoP
z`V!G0oyMo!F2}Yy!qqQ;2P9vKFKj8%5uchzI1l2*3KR*7gx)`#L_316r4c=42C8ZS
z9|9o-?#6);v6SaVSk(j|592c2oKB20*Jkq`S<;05e(qZQ4U<49}kx>DZK$kvqU5_A7Q-7F-#-i5(Saf5x{?OyYY*1EYUEh0;ivD&yFI$0rOVIE}bKh5^8I?7F2DiU`ZK6qC>Ysq_@oIs-QrbqM$p)iOrg1
zL3i_jz`^1uBh)z1%4jn2$9ILoqnfktS(VlJF+dCI?+nj4$j8g}(a>K40ECo}g`F2+
zTiFrb;~^rNzY4f;8L(_L*KsAuFSJj)sF6<(<2!#^@$)_5;>|G8TV3U@_AXI$9b-=6wZcm+)mL74J@UC__JyvrwnzB&v?YwmN-GM(iukERT~
zT}alZcL$*9!-~iW1WJ8^XSmrn^Cn+{Q`p|d5
zE4Y7dpNFCliWXyBWvDuL?`k`NVf|o4y=pZFAq5e81bL~otTQwAbH22lA(c2uuDn-C
zNzzd@$0vxP2rnSH&9t(CzHCq|RM}V#ufpeD?3jVjDj}J=C)e~0ht{a98(nW9gvm1M
zh8JTqu|q9iyv+6*jAOu7(VlgXpG^)jT3LU!$+z$a0A>{&+$bg6_gwub$vqXeJH}t1
zM6(wP>iZ9Du;Z2_SeO-Ay_`cAC5}MD3==4fXIbh?h4S(Vf5KBeb09~4yt&mMI){W)
za)S3XFcP1Nq03!<-AI2X`;3`&gR`(yV3~-It$g6LOj5gU5{ePwfj<}CgpUA^&Rl=u
z_Qp+f3(ljjchekZ*|0I5+CSe`(*RHT7CQ#YNigU~12Gp
zhKrDGC}B{LRCuZUO9A)oUvrCuIE^J&3|&bweHe_uS_g(b-i@<0YcpqhRVx?>+*;`V
zzHc{w48l4ku>`P;WHZeAz3{?sY>R(4;gi~n5zB^Vr@^lvl%?7h^W|{)vmHuu1qyg|
zqf2P$7yoQ60Dq>qz*Z`;O)#Vx;Kti>UMEpsH>K0rgqbyo4HA08S4
zh*SUCQ`J^xQuN7H;M9Ub`r>`chjsywV58cO*I(O#8~b8>X)X}(8U6OR&|!ZDl4+3$
zMYL0-kyeMJqo_P*;tlG01YKtiK3?PyM<{@AKguCF!F>oWZ0)PxW}r-+b2g3#x%&e{e73QphQi;R}B`SNP+}j<2Rf
z@aEOIFDP7sm>nvR4BDON>X%5~V4?h?zLW>ISGgK|sigI_66wJf
zpJHXjML&;STJ{8x68){OWQFJ~?UA0fs|qYy&tkpCkw@=XHd`}7B?#&t9nb!JDHi2~
zaSGPUecUM3rYn-*DHVT?7A|EMznR;GD>s)Np14)Ys_};RYA%uxxf^vh%e%X3nG`91Z%u*|H
zK_=Va)H@Zvia*BkXx}7bA}=UfpBnJfakX-Hev{Izt98SOa*}^Lh>RHefVBRoq6kzbe!IWQ#^tn-pQ!Y0Y8&^4A?y`YLV~_%bIg0A%MtT>XUcZny!dsEo-*n~J7vMRm
zBtntxV6GY-gw@zp?f<%W&IhLskse}%fyDoTGlR&8Y21I%{L&d<*I^T9R!VjGuUsf-
zOqlcq&_^2#@KdF}lyeRxFpGX?OMy;jmcJnbz9V{EePiittkQNUd(Rvz6FZFfq!jxkADJms
zWD|rPXOw@I@Vh5i0+Txp!EC4(1#Dg7i24X=86=6iJdRCe@QN)+_hBk=56r+j
zbAU>s2#}MPLHh6z83$Wy)&!YC=cri#R}RoZa9`;NP@#1L=C_$h?7;@M3*7C+xsHJI
z!)xvU=@R6cTSbVb>j?MGrc8hC;UosKk$Y0d4TV4$I@t}1Ul1O}
z3}2<#bqBvH5X|l*?|1=
zJKC|YrsA2g@thx!G+aZ_e(=xns)n-g3bFbU2Dz0ke95J)ZHVs>f2VS1L01?nkEM6A
zwL61CD^YN12;#=#mc7)7;|ihu2c65F(RY8Bl$|0PaD`$xvhrYNu7V6`6vLRtxR+2x
z#L-GMS?9X~&YesIs9HYCT7QRmdKwB^8dMD)|1?F^B!&`$Jfx5RF<)ykZp%6RWx#G%
z{y?OWj~@Gj0F{^fyX1DWZ@GEb0eFyGDq<347|+c1KT7U4^@J0xNO8lcIvfwpxki6=
zH|=^RNl*)g{w%duuX5$?@K2^{TT`@Q8I`V*Kl&!l1N{c#vJdnSToICqHNA@^a7Ldg_(U_L^r`5VcW;_98WUt2(-I
z2jFgayK;&-6X<7Y`9&%A$Q&^TH4uNb%u6mvk5?tguoHpVxJ$l8qPf5d>i05Vk3FKX
zrkXL<>Tw7hxS!OnSd|1k_4f+&8eu`*%VJQ3W`Z&$pb$gE&$om$ff316LS7(I5mANV
zXec5b;7&fGF%)Elyd3WdG2MyR@l(F|gM}o^lSX2h>$c8{CFSu$jcn-Pk1>B*0nfcW
zG>UqyeSo3-{&*Dyo_Hh}A_SeVmD(DalVF?Llqj@;?>(oH`?dC8p;%9CekArl$5%^0
z6QW;L9W{&r<)app!@g2*M6stu}MFO1zpFv$tccUkC+sTmclWDX-HJ+YaAv22$
zec%>Aq=H(5{&ZH_k7~Cu&=h}5=?{!b27voy-7%^npGJFM#jxLeJ9q}a2vFK>GAb(f
z3(g&|LdEmfh_hSRe2JcvlM6Q|2+1B0S^7{v-W%=IWwi?3W87Qv+FQ{@-fIW&)*
zPPsu?sfWozD@XE?oT}Q1>zhv>2|bB0Rbxw1iJ4>av-MIb%QP_Tw{(B;MBdOmTLwZ+
za=B|2huoBMM;}ohkP6P8d237j&Uo}N<|)NfIWKcAmnuvKg6ep7_!HX+@aaCA-&Wd*
zmx1IR!Le>+d>?llc2`2HCdZJMy(j(osLSD)Y3%@v0!R@rOaZFPZ8rfyw>CG)dQp~Y
zh;f{~MBhQ0mwBfA?kyf@
zIQ6g81)%&QBvSP&?3}arx@9atX!e#khxbiM!@mz#UFfW5UFG0rv+3A<-qKU3F%8Q7%%034$NJ|tird4&jt>au?eJ)+keGff@mj=C&4
zD{sE+zii{<+0zr=j{{3YLp^qe4>M}>s<@Z*aZ~EjfhW9`9LKHOuI9~Jb
ze}rhk@L~|B--^Tfr>^is3(6z}7`dKXHg>Nsp1ZU~#us`t$
zu}e8lbjyH2bXZ|kyvm0O{bZKML!un>)$XV9NZ#yf>`tHGgjB|{xcw!VrB=P0Cr(+W
zn&N^L(a?WWt>3gj);CG26cTqhb*zm
zUUGj*bLxZw@{%W|SxzSW;=eb^-(G%`OUd>>&XzZg_H`eN_5UB_
zi?q+@#GHY{P{vXc?-Yrm+N)N4GRMu0+^EH2&n;0P0a`6pSXpmW$@PeNsx6v@P7!=!
z*;hZ{sP6QobY2Tty2KHLjpc(@UeZWTc#3~n%%N`Op7M|IP{jAu`H6f07++A$sNgmUvd1l&qwq_T}o(Q|L7Ipxbox1UBZbb#9mMC<5o)b=~^T)A$AoBH0i>B~m
z8*YZd=A@k|exw8P1s1B5W^UcHV#?~6cd}gblN@d7IVGrKG&NTlX%!}ReOoYUG}V7s
zVgDqaZKj3Djd0l|_b(_i4v;9GiDaLJGkf@9hwHRjg>7M`Yu4f%J*g+F>#9|Fbg5N=
zhrLcI|F@_oiEMgBt@a)tJ=q@K-M$idM=L0|&1j&sGVp@$RFSw+)Q$m_DJhv?!A-2=
zi9CGreRu#VD8DQ_KK!fT;eF;6^Q(UlC%A9*E>3IPSS4ZhI=1+Srh)fjK2R>5Y~#=C
z_2Bd{usSptdFeG>T&7P^!+;{o;5ZBxNu|m|7;EB;Ar#<@A%$pQkUP9gTL!+F&e-UH^_{&nmMamVppzdW_Ga4@y5;qQY|LPDPV`oO=*sq
z>3LY9qiziDg;8SBhcSO0150J~G>u7H6A!LbZ?=G7u8^qM)^YT6EJC&4&Jk4gPur&Jg|wK)2A48Al7X6V$!jJugY
z>U0d1xm~VWU%J>PbfM2DxDz2l>yWRd5-)qSI@+wXF<50S>g2eMEyk|fKIi;TCKZLq
zP&<{zzV~^$MM19);~{@TeOj_x&d+^IwhLeLNFRe=tv0WJe`*k;JPljOQEo()j9Zl4&~h6p_jt29G&fv;
z$dS-17%+mFFtQav49X#ZtSA^)T9MI}m^%cw=y#*RB*-F87%YF7eh}{U1Pj@PtgxxH
zL0#BhL$r78ZcqKlqO62>*WM_yP{0N8)0*bEi+AQLpMxe9)SH`!*97vuzC{PtfEco2c|d9KKCpm1b6|`r2VmE{{hJ5|YeCVFetsSzBvvN2
zEDP(b_I7NlMpyiodWGBgwT@%?Iy$)s+(V!k>O!27TF8F{5sVRvehm~^-!s))4_QSI
z@7^DE4wGFr<|)TDV5`8l;Wmj%N|&w&V-+-qBy(9k7p?ni7lk^kRCvt#!9?1(_pX|@
zd;o(|upmo^5;OqEz-W$6NBAPp7-twgSIfjU$KZ@mrO`>B9Q;=`ppm@RX={Hm0u&NGYW*W;^$`he7
z5hdX@kt7zX7pn`$7z#ES<29QT&WY~#l3pbWN!fpjYY=T^Ym_wruTFr=CP!%quV|pd
zjdS##5Q^7)ktw#bSprA88^OzIYmK?o8GL!14pX(lwRM~DG$M?q$B1q7-APIyM_k}<
z1BV>GtMW3@3yvd~vp7gahH?Q)z}unG^QM=|gyqPn;s-_gkh
z7m&DeGZw6Zot9+@-`t5md7zgPZ%KFsS@UgFMWQQ26I{dIjHY8&y(VpKwh(EMXXp5Gj8M
z%KNE!Uy{n~#|ckKvO9T}CUE;LouTvslRVaiijL|Z!=B)bS*6E9{LvF;ZQq>;?Pd+j
ztCKysb7|6vx2re4(fkXd?WY*UzlzF)&}HgYbtzuVW8e{0B)D3vw#Dk
z)8dCDGgmb1AbLFPj+09HDA(CZl5Krd%Mg)Nn9xCHcmDWMzPhjR#K+m$GKa2Mn!rf*6VNNk
zM(#mFa)UN&NBJ0TOrM1SaHQu}H`+wM;Ou}e2WsEW<;Ig@BS@5%j>m9n#PeH{=q`A^
zmHgv_zQ|hHv0p$P4T@}oV1Iw+AqRU}|B6yj)YzznAQ%AMPIFu6^IXaAIkKMikt@ya
z@ej!5HM;|=3-20R^7xSY76hiNUN#y#3Fa;~s~Vutoh1kPnBr=XMskHHPdOz`<=<@)
zFGaCV6TGA0Hs@=^$BMbAvkUoxqYsyWC(NmUq3d{A3Gw)@XE6TFn)ZJPaQ#AMA#SuB
zgH4jOaKFt7PohATe*Agr2_%Ds8P0}lYYVOf;Vv+tmIyA$+t@Q5^ftxo!x`u5W)4Pw
z8hc~)i4jrTIvXq3$bpRHb77&H)nj7nnDS@ng%X^f6C__kTBZvSg{>GFYS${y-4es&
zq04$vdoFt9^7?sVZX18u4mkHLxLHx_9SkGX$@*1^7`CLRIV9H`3s+l*JO+s<`43h`
z(8|9E=q#Nz_KMWh?}3kM`M0emtg>cMaTJY-w&oxp=g{~bXvUoeTvXQGpb0SXnO8+%
zLD@gJ4XrR0TN>}@lcuHZ*?GQYFX&QyP8APZY*S39*O;P_@-
zv1_+dq7@_f{xhWP_z7=yYkWB8q1hrK&zS1#h+iT%l0;9t)bZp2_?Mgz
z4Ee2F>Q*MStlnRVOAx|`wK^;$9Jplug-E{=b&!#z?6SE%7~D`gBxyzx=?jjG
zk?v9R)7Fm`fA@c5+R4SbB^tVlMRGnNN_w1ReKP40jNoKFDb1a45RYSzmf#T!`NuaD
zAAJ#_)KQ0AON|{lR$3u;!t62LB|G&x+<>amZ*E6^yCo|)FH_$mf@}T&>Oth}TG`0q
zY=qlvYjQTxWERqa`{bg#$H~xOwij~ZxY>GN#AUlwC&qst-M@3;RYbw|$OkPu^NKE;
zw;>_PEz7-9w=#%tMOQ?S?zGEyMmb`l5%p_>RmHC&a~_7)dv!~VD8=^iz0RgB=h^+S
z*D^x5tv$-6_$7=K7nvYAN~jdQ^HYy9@Bwyb>eX}#%{F9Ltzt@wat*HO0b9aZOm7))
z*Ob@;*JgjhcCJ+HKqne8y|{dGAen3Nk(5cArU{xq`6OZUlnzHrf84fz`gGmI_!0dV
z5%e75H5PgDU~NMUpLgOi4ymYXWvYG)Ike1fb%?1g(V-SU`HOfvnM{g6_kIzkcF>E)
zpMtp=+r_g`UM!9{y^uMVUhpGi`_wXWF-gw
zEK;8*Ru5zQyqu}qSXHTnx4a4`W-9X>>z&sR8mh*c~<
z&0>F9|7(z`V7r3KMq|f0zU`@vN=vr3UTNY?^t2}O^fVOtD@ZQ&OjW^6D_2@lD4;=F
zZA`>{i2?&Bj7wIc=ecAPaAN)xezZVNM}=jbi81Od@n#%L{IQy|oT1R;XX0G4!Zq>ZPJ^3iSe?XW%leuZE<~jp7>}G-d^UpfM-SM#PY?O8gd*vvMB5)1x
z*+{gZ?&C%pD=PgW-JA9~9qxc9s62@dyW)4D)L%h%7VBoTMmj-JgE+HVyvUaB^*(=h
z(Fa!+7}f{NttPt97;LR^W=}}*jP0~>dLhA85Q-FGTO+8QK5*MPQ3{{x-NO#0Y>Yh#
zGR2q!=kyS+WAKxD|GC6N0vZKR+k?f3uCzWW6Lx9L$X_;cMX>%l`W2kWk!{09d<>|7
zNlXE_lo8IjtM67tJjMPmPxi|e&JBNIt8JGoI@WqFZ{#Kz7Bx|3Gun(2&?^P8KS}lg
zM=O?wH*ZfYDXYJH9CWi=^kxl4w*?bYkaAx~SW*qA#~a}fYdN=9`Y6vbH&3L!>>_I}
za~nlew;H}-9@*G{55HyZ6SVXbs5_@Gq)@LHeu3Lp_61-TS!`
z2Fi;a?ZjS83Nx2sND{5!pj`@!)6`tswx=IUaB18#*Ko
zINvGV!(HqwH34z*wTkSf7^r_46X*Kn{}q>p(wMlFNZ#)GOnG5d;$*&bTxN6-h4Lq0
zl5SKb)1yllpnml0btKF}J1uN<&SYeCEbW)s3@eveG~CJ=4Xyj_w6HwYoZjvjyh*Hq
zPcwTfLBmr`{=A1Ml`PGb?kA4B8WywI*&m_W1C(7vHWKe0XI)3A+X#PkT>z3~H7lR9
z*(_2{{%3WZxn0D!;cPeUPN56Ws=_^%w5|F-drR0*(mO3}mR=v=*=q}O+rMW)Wd%_%
zT4igAP(puX2H2^o=530}nm&|s8aPP_Bn-xY$xkl;ONIrJW24lzMU&3)d>yScQ}XrypjsHJFfsH6s{r~
zG<(~c7_ei2`(%%Qo+;w@A2OM)B;Bx8?jRY;*+58pTt)x}1E#h)PyJtbaUuDx>sc^<
z2@%vkdyuil?`MD9_qxwyfp9;XvvoY^N6%)Kb-noBRcQkU%ya6e62=D)H{ztX&PaVN
zfQ05jy`$V2Nbz$-Sgjekl6Zs*Y+n-dLr!r@g)#M&T3%SEVU$YD96p5JEBH5M2mb2k
zy}HDKvZgR3Fxh$;Q(8)}=nh`EQPgxVjJcq)n+7*25l4RxUCmtAR9v9?*2TjPP|?q;
z4n(7kI*<*>&k{EtcdpGPY`y<(Os+mUyW^x))$>gu*qS~MfhziXbZwlmp5+6r(mDi8
zG9*?$zM-e+D(~504KOC@VUBc1lZ+PLIv5!AyB1IE_4k%&7Es{au?`4O#zVaB
zJ~eg%t(uCb-h9>cql_wuhl93IgJv^lCposo)lQC_%)AA@qn`J+y%zHhUnf`
zg|0NsRig(#DxKu0x_J0TsBQl0Ol_ibpgAZeK@DV!crbuz9~=WD5OV*~wrSQbUS#r-
zLexqmPz7o+3ctC?0!lLZ!++_I&1Ye#6!_}0=H;TY3T%vMM649dN1=ti@5y_+66d3N
zE+~JG6fx8-ZRh^6l5L^N(~tF?Ma^Jyi){XTi);_~+V6$ezTdEe9q*sX^1n=v41rQ3
znAnwpwwBCkv=?c?JXUVT|ru;n)X^mC-ukC+;HwsgBme*tUwEL_1%
zpvlBwnGb2EbQ&5m9s^pVTSxK|QG4UFu}Ku7gm=3lFWL&}g=8lk81_wNi`b;KD!7h1
zabOGxm9mC$Jq-Ee&QH8huXQ9Dp(K`wSD$A1?u3oS6Sut?x3@pHGZ__U
zel3c*h+HXj>&3tY%Yhnk4nh){(q3Sa&?%$b43SkAXAnIiV*K6!s)$!igkNJcp&@U2
z&28!UO4Tk~3lw}xr9$M9KSvRNT9tn=WwvJ7w(Xl4KfcW;>qLK^Hl$!11C>^-oNkWT
zez`bN)P*7KWAw2}g9*XAZ$=JL7EnwcNe{}}ji8h@`7{SS_pQysjREUoh4_;5KW1~`w^virmfo$RdWId
z^4MeK9(hv%Ng(H^ss`J*WJqk~gy|XkcaG#1mC`h%ks2s1m@hVO@(V-rgp;$i^k`@<
zu_i%yMSHnXQL^DXF~+11Vqkv>ed=Ci+kQX_t(gC03uzB)tx_N+Uk_`uzSMKff%7wr
zmUWsm&kPvLmnhQmrAX=PL6)5~hEH{`q87iR`DxWak5AymOG{{ICn{gxv7$2}!5j|9
zMz*EXiHJM9aWPOrI2d4H8Xm|k?}N-on%*xC-aRSxFCxlIev`sLL0f-5kwP2-_%T}9
zD?n?Yqe@(zS4J=Z+%V`c`QG2m0JW_s&a#v*L_0=2dEq
zrSYSR(}I574)x9p7chTa_(sZK&h`1PEFqRRfZzG1_MogI6~mmLAT7Q9@Q}4aWrQhn
zfv7aP1^7TdNnq$V*klT}X%_zR({XamhvL+LZy3T9c}QLZL-y0}oC7(pOBz3^Ao&25
zUjg#W!_SMo>BLF5Gu$MUK8Q#YipX5yX6L@EGZ85-ne_*4xfHhr!!
z1G*j9KJ_F8T=K0&T1qAWg+?JBCW0-xC%_Q!SsGUW6;H94=2hAK4ern31#5Qz^mI*x_$XL8ykl%l%?3JX}1OQX4G
zC@-%o`C_r_Q1E{kFloH@paycpTN(uYZtn8$@o+r!98km51I=R-$KH#P@v2brlYk?)
zv1?x!Kk1lJTSpV`c-!MCs>j?h7=4y`^*E^{7%|(*QO3iE#rx#9luMBjs9bgWC%|Wn
zO7!N^M;j+I3?No71V&{6nCgK!={6#f^erH}Ti9}n-lcypn~;M9k(t9Q{o&~Vl-z&P
z^cSL8kGll#ck@W7@SWGe29{Bx`0^F4C!okwrU?Q^(FMUuWWQC@Vuy<
z6?VYnvcg&d0WY8W}vx}F&ycob*qV<0?DGo5Cp!)7X
zMLmulJAMn6=W)`HUSr2h6o;orz+y|*$aa4(xB`E>SKsloqu1yp+w*op3)6}?q)-?-
zDCX8qj-jBv(=N(1x3rONZ=s9F5SYass8@3H+Db|yVAWi885wL@$}JPJn@hRG1Hvm1
zaODV3Lwv6CU6TKd8y-J_qg>eOqT>8q<3xb|d
z*i?T;b^=_xvVk)Z6}(6}iyKdupD%b<@^T@jkA~UIq!7cHZ-Nuol=e;kd|B0~cWvQy
z{S{$mnt6Ea1P=B61%R8-EY#FIL_Bae0yShAc*rK8fDO6J*P9Ke$(iDO
zWE8lWyl^F3p&6z23!ZqA)#5OPr=`FM6H!D*UDt10?pK5QBbpOGcXWvM5aYFlr6M_(2b
z$|B#(oP2m(d8tkV)N
zwd%}WdHeHSq?9+R5B^8v#4Gt+2(`iW*}c_3hDP;I-DgEy@cz1+Y!AcfAD(}b^i|ZJ
zfDQbdH4`9Y;
z_@z_L{4%{cjObzxWKC4kO>8lB6;}NPF_5mX4D_-)6;Tz(?NyMzKxGY#GS=+O`(wZB
zBfr`fBRPn4b2L*P0QIr)*u;O3CjS%pwbiK*5Vf1Z(TkSxguKR?d(n1tPAMaNA^
zY9y;JGsVI5`O2u~@WpHr{DjREWya%MWXo}pCzrg>3L2>rxEC?KeH!gxoDD#<++AgJ
zeXMnv-e%eXL>T#5y=9V?u!kRu8Y+GKo^i`Zl-Wn-=<_C$TvW&^`zwEh3qMX&x(j&U
z%BT0oC&x!h44(6%Ox})-u2uCQ^!vmzJW0^#KWEPjRc6^5b^}>9^i=$4yLegWRuHf9
zB1e`p?#0P}hj(c7m;+=J`6neyaIRDHc7##ZOOWG$6eB1_>)SOSx4KhN44B{aUuer=
zBGZysP-xR%b;}!NXt00Nh8j)Q*@=d<87+LR+zct_}DMvz71CcH+)-ZpbxrXRXp8s<-0DRfr_lRmLm{d
zoUwDv{KQWSa{(+jFSWM$6n?jKG8t+XP)@h6e6hAX;wfo4+D(u(fPMpp)k}F#Ml6xt
zDwFZ@BqqE-hi-pWc$y80`lWE7=(~6{l+Ugvq{ndxo&(y6=_v6Y>(aoa18BizPGaAw
z3?!W1`D?=jkKGPTuAIMJyG6kp00_HE@1_b1A&TVZ+0@I$vLb!;SobC
zZ&taC;yr(+Ubxp}&sm;fL9-z8#kZ5f3+df~-$XN|g`HJU96+FgQFW%?YO`A8~&gS7o`{_x@5dQ3#$z@>8G_hy6Hpo@
z&MwPcvTHGBQQ;x>s_Zpyn4-aZxeYvA2#z?N2D;=#Yy$vMH?!m$|Qku6SsT@Z5-+u{`^gXxFFO;+9M4{W?=W4
zYzGav?0x5HZ|pgN`$GVTphrXnt5cr|luYoil&Y4Hmm1uEdCG)I8M0<&S=g3*BF|OHs=Hu{sV1e(gg~C#Bk3L~g^{(sdF{_*u+^em^4Tj;HqA
z=4&VxO2T7K5ecnZ>5xQc8{A5`{r4(BVz9x7-LUe~j0ThOa|MIYlna7q!bC(1)XxiV
z^!-Zvl?i+pNWH|*B$;+-kXKAvp98YhTl(QZ67^jdug*?Th>h1Qzr?$G{&BRfcgSd<
z=reN^4SiW;;2q{PYJC8le53R-qRVwY>QQ)Pl|wign^I#uaZ>PrSw=hG+n%@W)|HxV
zGc(_cGAwCNfZ@3bqZ*;b^_Eh>fJoV9$xWw2j^dVT-J5`evs(6V;wsK5a(VJ?oABp1
zE%4shoIY8A8Dicv630-4oT(FiV_@$WE<2+rb=n~_b~;&Y=~z?qMJp#jzi}F7q@6uM
z0%l^JXlA&$vyJ*1F!kaN+*zzB_eQj^xZ68UJj_wm#7fhNT+(fLC3kZ`}yh
zSmcUkbfeI~`shAyXgqy6{$&&hE@F`+vm@@LgBU#Xo0dIy}{pa_Z{I2jI9-L?$IzEF)z=9mjXY?9R
zk$pD)Ax09H2*d?ay7-^dICCklc}
z9k;zX>Kj6Q75bQ_=yi?;(X!r>lj5W7KaGrb>cy)soha#A9c(J+ioXBWzFhlEls%o)
zRNh$gi3}mCJIx?^`OwAULk^()dH-(jj>+`bd8ET_JzOChUrD-s@6^Eg`Is1EuIn>^
zP_968X^t{V;1+>x{vI)mkHXYqD*RJFh5<%)k+Q&v0nHUO=x0{CgOSueJsn`H&HhwmY6-#Hrv<$3-nUvBh7zMDmK!LNcR
z{1Hv;2y6r0t3OjzM3$1#@-qRgZ>342gx=AvrAk|onVX(HUF*#SivP>i%=7mk$6)Vs
z%#uLpYp)UAen?I^$PxJ&zNuBZyE#aeI5Vb**OwfiRpSBuh95$|6T21bA3X&~QU-Xn
z(IY(7SyWpC<|}l%W2d$7oxScz4%I8En0Vdbc{$`K1?vQ3Tt8^y-R+P3BVv4usN^<&
z9r!x&zLTs%J^m)1BHKl|R1S#1S)Hus)D1*LuJ_8;8p5n?1y1Ko4t}zG>kb7|@eGg{
zWS)-VAP-dfjI?3UX_#qsWS9Kpb!dK-vYE#W6ePLF(QcUfobfv%?P28aD*`PjDnU8}
zG|Slbr3(AC;+LCg>B99SE5!_}IM}zS&`Rr%aa!@=OX}SnzLwKf0Tp;8P*sCxI#z##
zKoWVXDSyaOSXo|F*xVhmM)ReF*39a7v9aY+<;ar{PmP-ib2p8YAO*vC7C90x9@924
ztDH<=mI&6r2I%m@}S42pA|6@7gie(W>(!(v*m(+?mABU4;YWcpLW{f-!ooe
z4vb?3^wk~1!`|Q%o~9jh;C&4aj3+J`yu#$HfwH8Fj~=13Xu3sSD*}5_5Kna;y{_oa
zT&=}C7%fjX=owuVs;#fOIf^aFf^I_htwu4E5oA6zqyY=3X4U4Ap|-&k-}|O?)6sc%
zq2k~??!S~h@TP+wp
z-wMfw1-AHTGW^&NzZ%~n*v3DN=bW5W9(jM_D0n#pH5=0R+miW`hh}`ZgtgX=hk{f`
za`EWD2k4kldrG1pOpNHHyI$kvL+TfaW)2gAKvWKqHlf&xEW@V9?m2L7zY@JF((>U=
zuXDMQuTCueG|0|z`Tj$QbpUNM(3%kTuX~e(nRG)A+8kPCIOKSc%ubUiXMIBpPa*yhM+MZf|P$O2nfD
z6dKTV8Q`j)nt3zLWzyMNC-(+MYYO*Fe0H%a85k+yGV&Gth>&km>Gk_+XaELLZsxhb3f4Sa7VmonC8%N2K5KZW}4asjdTz_HVY4C$cUU$uJZ@C
zd@Pv*nB00mFQFc;C!~?EDrus8;(-HYi0D%ZmFR
zA?&lMBhNVbEr^*sVYGtJH7P~~c&_KTSU)<=-bh<|XNB3Dam7(|zBL8Z-G*x=@*j$t
z;cDIN3ayNgK!mU#z1K_0gyGA4aM733}nlZ-Dl1`1W=>a5p&D$(L1-tH#AKDAW5T&3r!O
zo@DR4DbIG^j`wHKx>K1#;J-byn>YjB6IHo7FULQsK$^q_3qk@{b(et9p
zZnA*Jr`pwzf;{A7s|%h1$wZ-pb4-Ab4`eJU5X%i8rA_@0*H}ls@l>MmE1vH=okHZ_
zH;B(isHvOwcv@PLK_`udp)NjFMlZd(d;
zxK|M(5-^
zCOXJ3G-_IBI$~;PQ3BrFKY$!^#snjaIh6Rx$}Q|t`-hJ0si!aDFAV^cm_aS7$DgEVywsmxSCzak!5TFJk50>$a*#a
z$bX!OnyL4ZGeLm=&KUoTY=a5lOXq(=j$q~D=H=$$<9^G5;
zp~5^|+(O)3FYW)kr0+fWE_%=Ft7f&R6VFygkjnUYwQGn(pfH%Ut!V4?
z3N<}?#kD%&t9~z=W(XQ%ky(M+cUDHx&{?Eb>?)Lu*L_8(wkWlc
zt--XwduhA<-Wif2TJe0=GPC?lQWp&RAsKwd3mIC+j5Hd_4+Apu!Jm;Z!9yWT#pM+1)YQ<(yV=5j0T5}c<4xe
zgy!`^j}e+sFs@ZMGFDF)7s4%A@?_Hu)I|ft<;48kf{T-7Mu43fK$W=9@P9X>5q7o93cUCAGFP^dErm}Id
ztavdgD)1))64COog0PwqVSju5hKGuhH&7?AdO>>MFdN!t&6&7k?n2vqL+G1T7&#$j
zvJf@*aY8E68CD*42yS$eVXZvyG0&ZPQf_ZNs%Ccm<2x-Ux}7Drpjo+~DK?Sg;P3x!
z=}2L+qM4R&%Cl%Kj-3HBrTEwmHhFFs^)~yExi-7w`=>alFXCVGgVhF#V%Ox^n;8Li
z9rr<$0NH9SpfmLj@1pZIuj(ruP$Vp91q~|3)&EO3eQR)5_u7{RMS`{a%Kt?S3#0aX
zptl=N3rl9m_HTZ=wqRQEkY`Ra8wQd}Mz(02D*<_UkB-+wxzEbvh+hGhiMg1a!THpO
zuh_JHZP9b%{}3v6vG214?aW1R)VHM~NI>w$C<x;35xNOKAHJi7xl>q@MrNjI@{R|%D9
zkZ}-0Ch(0+V4|IY8mym;c3{qEa1R$d6U*D68Lb~HiAkG4Iaz$tE`P94?a5y2(zt*a
z^iG8J6~ZFbs2EJ6X-6MCy;le-uJy*=Ci-nrn`y&7|1W*TosXRU>;h;Ys_T=a(66mW
zj{eTvh1tp&$e=lbqJ#(YTP)B7anI1@Hjw0wE~i$GX5JD~Hl~IdIk_@d6CN5&iYnq}
z7GO#lO>^r)c7L;hqS9)uu;uu}>@}I9tZayB5#3e^{VPS1nWfw(N;D@dMkysh@|UPn
zmj4M3&S;5CEZLz$*XBS3C4GoK7-dc@+j>9erL$TH5v5CFIt*^0K^!S$GIm$A
zn<yR?j}d{wai2^*Ai_Jl4FuNb-+#p@bMGX2u{4+NX}p$^(q>cxs3eja3*}KI
zD8L|k4`2dnnH67JB@a~wkTb8Zc)&Aa-EyxqHY^g5SR7gSmRVv(cvpWYJkg?yfJZaA
zM#V#rZ5D=KDY(YJg|F+papA?4l7Ha=GH9!&+^M+#rHt=+yw!(Nd6wx|aSordvT=Y~riGm4pF=Q2p8BcMdOQPwCGF
z#A|YkyVyoMqwr&kj$UL4rxN4%Nxlz>Di>b4&@`4!%Fj*bcSXp)g_fjnY?VfK;jDSS
z0T_%j7tQ7s*iO7tla;%Vb?ud{yjARpwQRi#HQZojSc*BvB=sJd@SE
zzq&rQU#~m9n)`4wdzd__{DCnX?d<}_%`|siHwaed=snGnr
zduotl0u@P8sI_hFT0b8Mn@)v$x~igb`KfzhJgz>SIjGfHMC9shkIHfNW~<)O5WPBU
zIZ@doIwO?W9KZE=3VH4~J)eKTme@9aO1-&7pp`N6AxvLaZ63;bAiY~~*_5Ky%^}F+rW{oODL6cOrpFfWwU={vSRb)3NKuO&g9OH@YZg
z-FNW}6%PS5iwJM;j%RX2jo(UTRkHY5Rq}-c|DKA=FRao%u{TSUlv&v&{_Hd*{4`B+
zCaGNYFccn!%!E!f(ZQ^HY}BNuv*ePUs+RP0xS%-VGQYTWsiI}T59C0e19fGZKcJ+{
zFV60W_jWb?(jg)0L?08IX|Js|7UYuIg@#v0jDGxOqj~rZe}y)}IhXp6pkQ3tmbD5c
z3z^E@o97;hi%Y)u>Z?YcrSlyRj&qb?d5-<+@vSuhDE6Fe{b1V+?-8z|;KZw~ObQ4G
zQxRFfwHzs2G(#ER6)`HStwjn1o9|))Nm3Vrf9W8X05A5g<{|Nr;vhQ6@AK>O4==bo?
zs9A(o)gRV2SuOSTc0yUgK!BA0kiXeUvCqGFZl2n-3y$ELpIFP!uho~K%o*?-bBXMc
zfv&o89~|7*e-$N?tUS}l&LN^?!Df0k_4YzkJHZ1VUiQCsSSaBMlH&M^6#2nsF#@8O
zzwl~Koeq{wD^@8)gKyZx1%9IATi5i9)o{?ZGQv6?z;ZRX#r@tMFA
z;(UwmFxNMfVuCw1|JzJW=fQ^m>;0CL;jI65+v8>ZE~3Ms+)r*qZ4wwB2lz|Z2vW`8
zzt7meVnTFE{U(n^D1A%IY%c1w?OWX|km&ec^*B6C@H4FI>6y*C%0#J6ujRhh#Q6HX
zXBH{dm(#4-ohGudDxP)|oJoqPOMH4J6nqqO|8wRSLz7AFXr(L8MQxd^s(dX;>Td*=jz&aE8_SQ&1OMo*Z3
z3`~>L|7d)8tuYhMmLpl!K31VPXCxJC&>ijxi_J7hQxWgE6Ucp={?Raa3=&otE#!WI
zY1LkztH*QX0EQ&;^0`^)w`>AzItx+sk`T@erI@osDzo2zGvS@!4;H36I#Ysd7|QlB
ze(WvMZ6@iJGvsaCdEx1uqgqr-tgwF~=8v8opl|@y?qSEUbcS$;;tfb~{v(~f6S2RW
zEqd5d@GG`ZcP_&~+g6n;TwS%Fg+<7nW)W19&l^%#GQvsuUzkErvjCSh+7Z*e{yuDxBVdNXf!hUm|(!pDXmeNPO4y3
z-w;^}(<5nSkWKdP>{(Fb7><4{(lg*~|NG6wfAnO^zx|l6S&{R+Y5swubFy7Ur*Vna
znA2gQSQjy0i>b_?Z<7okVY>Whb)}gA!L)emFOWe4ktB6dIF2w*Nt&Ehxj4TQk=Fl|
z5E^>YNoAb!(ANMLDMz=v!pCv_we2Qx@{@%75)sE6Z~VVGONc)`8w-U7Az%#T)U@}r
z1`EA#k1p7d8_Wylei_qnb@c=bLBL#K4Irn2i;XK-i0{7!6l^F8lZMDZc%@+iGJ-NN
zDPE|ckSt74NRXFDmLKv$Rs!PS|J&td6(!LBkQNWG5ct2O#W32e5n_%fyAdD(?LlR|
z%|f6sjtJRey{_t;J4ZiarvM3a3{Listx|!=**(V=MLn7pjCt$V%+ih`0IyLsZ+!*c
zC8X)k{X)WA2xk`zMq%!wI+8@8&&SuO0m-=m^#=xc&?MGSe@$Sq*Wgx8psyq(Zx44(
z;!sXtMkGWakUgbP119LApQ*3DqyMPEk*UG#<3e`i0>WO0zRw6!$OuWQ@YzZ)wZ@TO8((lF9$TkibpyPB
m**tKKPm|WTSz@_}I{SYmCfyYW1sfzND8vtBV1TL00{