update nys vs nystrom experiment

main/experiments/build_command_line_array_nys_vs_deepstrom.py

+import numpy as np
+
+if __name__ == '__main__':
+    batch_size = np.logspace(3, 9, dtype=int, base=2, num=5)
+    num_epochs = np.linspace(200, 200, num=1)
+    gamma = 0.003
+    subsample_size = np.logspace(3, 9, dtype=int, base=2, num=5)
+    networks_types = ["nystroem", "deepstrom"]
+    for network in networks_types:
+        s_network = "--" + network + " "
+        s_gamma = s_network + "-G " + str(gamma) + " "
+        for nys_size in subsample_size:
+            s_nys = s_gamma + "-m " + str(nys_size) + " "
+            if network == "deepstrom":
+                for num_epoch in num_epochs:
+                    s_epoch = s_nys + "-e" + " " + str(int(num_epoch)) + " "
+                    for b_size in batch_size:
+                        s_batch = s_epoch + "-s " + str(b_size) + " "
+                        print(s_batch)
+            else:
+                print(s_nys)

main/experiments/nystrom_vs_deepstrom.py

@@ -9,10 +9,10 @@ Options:
     -h --help                                               Show this screen.
     --nystroem                                              Run the nystroem version.
     --deepstrom                                             Run the deepstrom version.
-    -e numepoch --num-epoch=numepoch                        The number of epoch. [default: 1]
-    -s batchsize --batch-size=batchsize                     The number of example in each batch [default: 50]
     -G gammavalue --gamma-nystrom=gammavalue                The gamma value used in nystrom.
     -m subsamplesize --subsample-size-nystrom=subsamplesize The subsample size for nystrom.
+    -e numepoch --num-epoch=numepoch                        The number of epoch. [default: 1]
+    -s batchsize --batch-size=batchsize                     The number of example in each batch [default: 50]
 """
 
 import tensorflow as tf
@@ -21,7 +21,7 @@ import skluc.mldatasets as dataset
 from sklearn.kernel_approximation import Nystroem
 from sklearn.svm import SVC
 
-from nystrom.nystrom_approx import nystrom_layer
+from skluc.nystrom.nystrom_approx import nystrom_layer
 from skluc.neural_networks import batch_generator, classification_mnist
 
 tf.logging.set_verbosity(tf.logging.ERROR)
@@ -29,10 +29,18 @@ tf.logging.set_verbosity(tf.logging.ERROR)
 import docopt
 
 
-def deepstrom_classif(X_train, Y_train, X_nystrom, batch_size,
-              num_epoch, dataset_cycling,
-              gamma, data_shape, output_dim, output_nystrom_layer,
-              X_test=None, Y_test=None):
+def deepstrom_classif(X_train,
+                      Y_train,
+                      X_nystrom,
+                      batch_size,
+                      num_epoch,
+                      dataset_cycling,
+                      gamma,
+                      data_shape,
+                      output_dim,
+                      output_nystrom_layer,
+                      X_test=None,
+                      Y_test=None):
 
     with tf.Graph().as_default():
 
@@ -77,7 +85,8 @@ def deepstrom_classif(X_train, Y_train, X_nystrom, batch_size,
                 # testing or predicting may not be wanted
                 accuracy = sess.run([accuracy_op], feed_dict={
                     x: X_test, y_: Y_test, keep_prob: 1.0})
-            print(accuracy)
+            lst_output = [str(accuracy[0]), str(x_nystrom.shape[0]), str(gamma), str(batch_size), str(num_epoch)]
+            print(",".join(lst_output))
 
 
 def nystroem_classif(X_train, Y_train, X_test, Y_test, subsample, gamma):
@@ -87,22 +96,26 @@ def nystroem_classif(X_train, Y_train, X_test, Y_test, subsample, gamma):
     X_test_transformed = nys.transform(X_test)
     clf = SVC(kernel="linear")
     clf.fit(X_train_transformed, Y_train)
-    print(clf.score(X_test_transformed, Y_test))
+    score = clf.score(X_test_transformed, Y_test)
+
+    lst_output = [str(score), str(len(subsample)), str(gamma)]
+    print(",".join(lst_output))
 
 
 if __name__ == "__main__":
     arguments = docopt.docopt(__doc__)
-    print(arguments)
+    # print(arguments)
     SUBSAMPLE_SIZE = int(arguments["--subsample-size-nystrom"])
     gamma = float(arguments["--gamma-nystrom"])
     nystroem = arguments["--nystroem"]
     deepstrom = arguments["--deepstrom"]
-    num_epoch = int(arguments["--num-epoch"])
+    num_epoch = int(float(arguments["--num-epoch"]))
     batch_size = int(arguments["--batch-size"])
 
     mnist = dataset.MnistDataset()
     mnist.load()
     mnist.normalize()
+    np.random.seed(0)
     indexes_nystrom = np.random.permutation(60000)[:SUBSAMPLE_SIZE]
 
     if nystroem: