Added SVM fix for bigrams+

classifier.py

@@ -149,7 +149,7 @@ def vectorToString(self, vec, cls = False):
         # granted, this is kind of silly
         # creates a string of the format "[class if point is labeled] feature1:value1 feature2:value2..."
         # where the only allowed features are the ones in restrictFeatures, if we're restricting the features
-        return ((str(cls) + " ") if cls else "") + " ".join([str(i) + ":" + str(vec[i]) 
+        return ((str(cls) + " ") if cls else "") + " ".join(["-".join(str(i).split()) + ":" + str(vec[i]) 
                                                                  for i in vec if (not self.restrictFeatures) or (i in self.restrictFeatures)]) + "\n"
 
     def addFeatureVector(self, point, cls, binary=False):

movie.py

@@ -115,10 +115,10 @@ def test(classif, n=1, train_size=500, mode='k', iterations=1, dataset='', limit
 if __name__ == "__main__":
     #test(classifier.BayesClassifier,n=[1],train_size=800,mode='k',
     #     iterations=3,dataset='position',limit=[16165],binary=True)
-    #test(classifier.LinearSVMClassifier,n=[2],train_size=800,mode='k',
-    #     iterations=3,dataset='default',limit=[16165],binary=True)
-    test(classifier.MaximumEntropyClassifier,n=[1],train_size=800,mode='k',
+    test(classifier.LinearSVMClassifier,n=[2],train_size=800,mode='k',
          iterations=3,dataset='default',limit=[16165],binary=True)
+    #test(classifier.MaximumEntropyClassifier,n=[1],train_size=800,mode='k',
+    #     iterations=3,dataset='default',limit=[16165],binary=True)
 
 # with train_size = 800, no shuffling, bayes classifier
 # [ns]      dataset         [limits]        binary  --> +results    -results