@@ -34,19 +34,19 @@ def abserr(predicted, target):
3434
3535
3636# Predict (probability) based on given parameters
37- def predict_proba (X ,Weights ):
37+ def predict_prob (X ,Weights ):
3838Z = af .matmul (X ,Weights )
3939return af .sigmoid (Z )
4040
4141
4242# Predict (log probability) based on given parameters
43- def predict_log_proba (X ,Weights ):
44- return af .log (predict_proba (X ,Weights ))
43+ def predict_log_prob (X ,Weights ):
44+ return af .log (predict_prob (X ,Weights ))
4545
4646
4747# Give most likely class based on given parameters
48- def predict (X ,Weights ):
49- probs = predict_proba (X ,Weights )
48+ def predict_class (X ,Weights ):
49+ probs = predict_prob (X ,Weights )
5050_ ,classes = af .imax (probs ,1 )
5151return classes
5252
@@ -66,7 +66,7 @@ def cost(Weights, X, Y, lambda_param=1.0):
6666lambdat [0 , :]= 0
6767
6868# Get the prediction
69- H = predict_proba (X ,Weights )
69+ H = predict_prob (X ,Weights )
7070
7171# Cost of misprediction
7272Jerr = - 1 * af .sum (Y * af .log (H )+ (1 - Y )* af .log (1 - H ),dim = 0 )
@@ -122,7 +122,7 @@ def benchmark_logistic_regression(train_feats, train_targets, test_feats):
122122t0 = time .time ()
123123iters = 100
124124for i in range (iters ):
125- test_outputs = predict (test_feats ,Weights )
125+ test_outputs = predict_prob (test_feats ,Weights )
126126af .eval (test_outputs )
127127sync ()
128128t1 = time .time ()
@@ -172,8 +172,8 @@ def logit_demo(console, perc):
172172af .sync ()
173173
174174# Predict the results
175- train_outputs = predict_proba (train_feats ,Weights )
176- test_outputs = predict_proba (test_feats ,Weights )
175+ train_outputs = predict_prob (train_feats ,Weights )
176+ test_outputs = predict_prob (test_feats ,Weights )
177177
178178print ('Accuracy on training data: {0:2.2f}' .format (accuracy (train_outputs ,train_targets )))
179179print ('Accuracy on testing data: {0:2.2f}' .format (accuracy (test_outputs ,test_targets )))
