third fixed production

clclcocoro · Sep 3, 2015 · c191d6a · c191d6a
1 parent 410a6c3
commit c191d6a
Show file tree

Hide file tree

Showing 15 changed files with 96,963 additions and 4,862 deletions.
diff --git a/create_model.py b/create_model.py
@@ -23,6 +23,7 @@
 from pybrain.tools.shortcuts     import buildNetwork
 from pybrain.supervised.trainers import BackpropTrainer
 from pybrain.datasets            import SupervisedDataSet
+from pybrain.structure.modules   import SigmoidLayer
 from sklearn.ensemble            import RandomForestClassifier 
 from sklearn import svm
 import feature
@@ -45,7 +46,7 @@ def create_NN_classifier(genes, positive_dataset, negative_dataset):
     ds = SupervisedDataSet(indim, 1)
     for i in xrange(len(train_labels)):
         ds.appendLinked(train_dataset[i], [train_labels[i]])
-    net = buildNetwork(indim, node_num, 1)
+    net = buildNetwork(indim, node_num, 1, outclass=SigmoidLayer, bias=True)
     trainer = BackpropTrainer(net, ds, learningrate=learning_rate)
     trainer.trainUntilConvergence(maxEpochs=maxEpochs_for_trainer)
     return net

diff --git a/cross_validation.py b/cross_validation.py
@@ -3,6 +3,7 @@
 from pybrain.tools.shortcuts     import buildNetwork
 from pybrain.supervised.trainers import BackpropTrainer
 from pybrain.datasets            import SupervisedDataSet
+from pybrain.structure.modules   import SigmoidLayer
 from sklearn.ensemble            import RandomForestClassifier 
 from sklearn import svm
 import numpy
@@ -15,7 +16,7 @@ class CrossValidation(object):
     """
     Gene Scale in GA has to be (0, n).  n is greater than 0.
     """
-    def __init__(self, bindres_file, pssms_file, log_file, method, fold=5, undersampling=True, shuffle=True, maxEpochs_for_trainer=5, geneScale=(0, 10)):
+    def __init__(self, bindres_file, pssms_file, log_file, method, fold=5, undersampling=True, shuffle=True, maxEpochs_for_trainer=10, geneScale=(0, 10)):
         if geneScale[0] != 0 or geneScale[1] <= geneScale[0]:
             raise ValueError("Gene Scale in GA has to be (0, n).  n is greater than 0.")
         if method != "neuralNetwork" and method != "randomForest" and method != "SVM":
@@ -117,7 +118,7 @@ def neuralNetwork_eval_func(self, chromosome):
             ds = SupervisedDataSet(indim, 1)
             for i in xrange(len(train_labels)):
                 ds.appendLinked(train_dataset[i], [train_labels[i]])
-            net = buildNetwork(indim, node_num, 1)
+            net = buildNetwork(indim, node_num, 1, outclass=SigmoidLayer, bias=True)
             trainer = BackpropTrainer(net, ds, learningrate=learning_rate)
             trainer.trainUntilConvergence(maxEpochs=self.maxEpochs_for_trainer)
             decision_values = [net.activate(test_dataset[i]) for i in xrange(len(test_labels))]

diff --git a/cross_validation.pyc b/cross_validation.pyc
diff --git a/output/SVM.pkl b/output/SVM.pkl
diff --git a/output/SVM_best_chromosome.tsv b/output/SVM_best_chromosome.tsv
@@ -1,2 +1,2 @@
 #method	cost	gamma	window_size	decision_value
-SVM	256	0.0078125	1	-0.94674091758
+SVM	1024	0.0078125	10	-0.53006984705
diff --git a/output/SVM_prediction_output.txt b/output/SVM_prediction_output.txt
diff --git a/output/log.txt b/output/log.txt
diff --git a/output/neuralNetwork.pkl b/output/neuralNetwork.pkl
diff --git a/output/neuralNetwork_best_chromosome.tsv b/output/neuralNetwork_best_chromosome.tsv
@@ -1,2 +1,2 @@
 #method	node_num	learning_rate	window_size	decision_value
-neuralNetwork	5	0.046	1	0.545129728059
+neuralNetwork	28	0.064	8	0.314267765707
diff --git a/output/neuralNetwork_prediction_output.txt b/output/neuralNetwork_prediction_output.txt
diff --git a/output/randomForest.pkl b/output/randomForest.pkl
diff --git a/output/randomForest_best_chromosome.tsv b/output/randomForest_best_chromosome.tsv
@@ -1,2 +1,2 @@
 #method	n_estimators	max_features	window_size	decision_value
-randomForest	281	24	3	0.0
+randomForest	1001	27	5	0.0
diff --git a/output/randomForest_prediction_output.txt b/output/randomForest_prediction_output.txt
diff --git a/test/bindingData.txt b/test/bindingData.txt
@@ -1,3 +1,3 @@
-http://purl.uniprot.org/uniprot/AAAAAA 1 22 43 64 85 106 127 148 169 190
-http://purl.uniprot.org/uniprot/BBBBBB 2 23 44 65 86 107 128 149 170 191
-http://purl.uniprot.org/uniprot/CCCCCC 3 24 45 66 87 108 129 150 171 192
+http://purl.uniprot.org/uniprot/AAAAAA 1 22 43 64 85 106 127 148 169 190 211 232 253 274 295 316 337 358 379 400 421 442 463 484
+http://purl.uniprot.org/uniprot/BBBBBB 2 23 44 65 86 107 128 149 170 191 212 233 254 275 296 317 338 359 380 401 422 443 464 485
+http://purl.uniprot.org/uniprot/CCCCCC 3 24 45 66 87 108 129 150 171 192 213 234 255 276 297 318 339 360 381 402 423 444 465 486
diff --git a/test/pssms_random_score.txt b/test/pssms_random_score.txt