Merge pull request #243 from Palashio/ramya-branch

Replace dicts in sklearn models  with LabelEncoders

libra/plotting/generate_plots.py

@@ -1,7 +1,7 @@
 import seaborn as sns
 import warnings
 from sklearn.metrics import roc_curve, auc
-from sklearn.preprocessing import label_binarize
+from sklearn.preprocessing import LabelBinarizer
 from numpy import interp
 import pandas as pd
 import sklearn
@@ -261,9 +261,11 @@ def plot_mc_roc(y_test, y_score, interpreter=None):
     lw = 2
     n_classes = len(np.unique(y_test))
     classes = pd.unique(y_test)
+    label_binarizer = LabelBinarizer()
+    label_binarizer.fit(np.concatenate((y_test, y_score)))
     if n_classes != 2:
-        y_test = label_binarize(y_test, classes=classes)
-        y_score = label_binarize(y_score, classes=classes)
+        y_test = label_binarizer.transform(y_test)
+        y_score = label_binarizer.transform(y_score)
     else:
         n_classes = 1
         y_test = y_test.reshape(-1, 1)
@@ -308,21 +310,15 @@ def plot_mc_roc(y_test, y_score, interpreter=None):
              color='navy', linestyle=':', linewidth=4)
 
     for i in range(n_classes):
-        if isinstance(interpreter, dict):
-            inverted_interpreter = dict(map(reversed, interpreter.items()))
-            plt.plot(fpr[i], tpr[i], lw=lw,
-                     label='ROC curve of class {0} (area = {1:0.2f})'
-                           ''.format(inverted_interpreter[i], roc_auc[i]))
-        else:
-            plt.plot(
-                fpr[i],
-                tpr[i],
-                lw=lw,
-                label='ROC curve of class {0} (area = {1:0.2f})'
-                      ''.format(
-                    interpreter.inverse_transform(
-                        [[i]])[0],
-                    roc_auc[i]))
+        plt.plot(
+            fpr[i],
+            tpr[i],
+            lw=lw,
+            label='ROC curve of class {0} (area = {1:0.2f})'
+                ''.format(
+                interpreter.inverse_transform(
+                [[label_binarizer.classes_[i]]])[0],
+                roc_auc[i]))
 
     plt.plot([0, 1], [0, 1], 'k--', lw=lw)
     plt.xlim([0.0, 1.0])
@@ -401,7 +397,9 @@ def analyze(client, model=None, save=True, save_model=False):
         if model in ['svm', 'nearest_neighbor',
                      'decision_tree', 'text_classification']:
             label_source = modeldict['interpreter']
-            labels = list(modeldict['interpreter'].keys())
+            labels = []
+            for num in np.unique(np.concatenate((real, preds))):
+                labels.append(label_source.inverse_transform([[num]])[0])
         else:
             label_source = enc
             labels = enc.classes_

libra/queries.py

@@ -147,16 +147,8 @@ def interpret(self, model, predictions):
         '''
         modeldict = self.models[model]
         if modeldict.get('interpreter'):
-            if isinstance(modeldict['interpreter'], dict):
-                inverted_interpreter = dict(
-                    map(reversed, modeldict['interpreter'].items()))
-                toRet = []
-                for each in predictions:
-                    toRet.append(inverted_interpreter[each])
-                predictions = toRet
-            else:
-                predictions = modeldict['interpreter'].inverse_transform(
-                    predictions)
+            predictions = modeldict['interpreter'].inverse_transform(
+                predictions)
         clearLog()
         return predictions
 

libra/query/classification_models.py

@@ -11,6 +11,7 @@
 from libra.plotting.generate_plots import (generate_clustering_plots)
 from colorama import Fore, Style
 import warnings
+import sklearn
 
 warnings.filterwarnings("ignore", category=UserWarning)
 
@@ -245,13 +246,11 @@ def train_svm(instruction,
     # Needed to make a custom label encoder due to train test split changes
     # Can still be inverse transformed, just a bit of extra work
     y_vals = np.unique(pd.concat([y['train'], y['test']], axis=0))
-    label_mappings = {}
-    logger("Labels being mapped to appropriate classes")
-    for i in range(len(y_vals)):
-        label_mappings[y_vals[i]] = i
+    label_mappings = sklearn.preprocessing.LabelEncoder()
+    label_mappings.fit(y_vals)
 
-    y_train = y_train.apply(lambda x: label_mappings[x]).values
-    y_test = y_test.apply(lambda x: label_mappings[x]).values
+    y_train = label_mappings.transform(y_train)
+    y_test = label_mappings.transform(y_test)
 
     # Fitting to SVM and storing in the model dictionary
     logger("Fitting Support Vector Machine")
@@ -328,11 +327,11 @@ def nearest_neighbors(instruction=None,
     num_classes = len(np.unique(y))
     # encodes the label dataset into 0's and 1's
     y_vals = np.unique(pd.concat([y['train'], y['test']], axis=0))
-    label_mappings = {}
-    for i in range(len(y_vals)):
-        label_mappings[y_vals[i]] = i
-    y_train = y_train.apply(lambda x: label_mappings[x]).values
-    y_test = y_test.apply(lambda x: label_mappings[x]).values
+    label_mappings = sklearn.preprocessing.LabelEncoder()
+    label_mappings.fit(y_vals)
+
+    y_train = label_mappings.transform(y_train)
+    y_test = label_mappings.transform(y_test)
     logger("Labels being mapped to appropriate classes")
     models = []
     scores = []
@@ -418,13 +417,11 @@ def decision_tree(instruction,
     # Needed to make a custom label encoder due to train test split changes
     # Can still be inverse transformed, just a bit of extra work
     y_vals = np.unique(pd.concat([y['train'], y['test']], axis=0))
-    label_mappings = {}
-    for i in range(len(y_vals)):
-        label_mappings[y_vals[i]] = i
+    label_mappings = sklearn.preprocessing.LabelEncoder()
+    label_mappings.fit(y_vals)
 
-    # Custom label encoder due to train test split
-    y_train = y_train.apply(lambda x: label_mappings[x]).values
-    y_test = y_test.apply(lambda x: label_mappings[x]).values
+    y_train = label_mappings.transform(y_train)
+    y_test = label_mappings.transform(y_test)
 
     logger("Labels being mapped to appropriate classes")
     num_classes = len(np.unique(y))

libra/query/feedforward_nn.py

@@ -358,6 +358,8 @@ def classification_ann(instruction,
     # print((" " * 2 * counter)+ tabulate(datax, headers=col_name, tablefmt='orgtbl'))
     losses.append(history.history[maximizer]
                   [len(history.history[maximizer]) - 1])
+    accuracies.append(history.history['val_accuracy']
+                  [len(history.history['val_accuracy']) - 1])
     # keeps running model and fit functions until the validation loss stops
     # decreasing
 
@@ -386,9 +388,9 @@ def classification_ann(instruction,
         datax = []
         values.append(str(len(model.layers)))
         values.append(
-            "| " + str(history.history['accuracy'][len(history.history['val_accuracy']) - 1]))
+            "| " + str(history.history['accuracy'][len(history.history['accuracy']) - 1]))
         values.append(
-            "| " + str(history.history['accuracy'][len(history.history['val_accuracy']) - 1]))
+            "| " + str(history.history['val_accuracy'][len(history.history['val_accuracy']) - 1]))
         datax.append(values)
         for row in datax:
             print((" " * 2 * counter) + "| " +