def get_model_accuracy_filter_feature_size(self,expression_file, ic50_file,feature_size,num_permutations,drug):
scikit_data,scikit_target = dfm.get_expression_scikit_data_target_for_drug(expression_file,ic50_file,drug,normalized=True,trimmed=True,threshold=None)
for i in range(0,num_permutations):
try:
shuffled_data,shuffled_target = dfm.shuffle_scikit_data_target(scikit_data,scikit_target)
accuracy = cv.cross_val_score_filter_feature_selection(self.model,cv.trim_X_num_features,feature_size,shuffled_data,shuffled_target,cv=5)
yield accuracy.mean()
except:
yield 0.0
def get_accuracy_and_runtime_vs_num_generations(expression_file,ic50_file,num_features,generation_range,num_permutations,num_threads):
drug = "SMAP"
scikit_data,scikit_target = dfm.get_expression_scikit_data_target_for_drug(expression_file,ic50_file,drug,normalized=True,trimmed=True,threshold=None)
p = Pool(num_threads)
scores = p.map(wrap, [(g,scikit_data,scikit_target,num_features,num_permutations) for g in generation_range])
scores = {g : scores[i] for i,g in enumerate(generation_range)}
print(scores)
return scores
def get_model_accuracy_RFE(self,expression_file,ic50_file,target_features,num_permutations,drug):
scikit_data,scikit_target = dfm.get_expression_scikit_data_target_for_drug(expression_file,ic50_file,drug,normalized=True,trimmed=True,threshold=None)
step_length = int(len(scikit_data.tolist()[0]) / 100) + 1
for i in xrange(0,num_permutations):
try:
shuffled_data,shuffled_target = dfm.shuffle_scikit_data_target(scikit_data,scikit_target)
selector = RFE(self.model,target_features,step=step_length)
yield cross_val_score(selector,shuffled_data,shuffled_target,cv=5).mean()
except:
yield 0.0
def get_cross_validation_time(self,expression_file, ic50_file,feature_size,num_permutations,drug):
scikit_data,scikit_target = dfm.get_expression_scikit_data_target_for_drug(expression_file,ic50_file,drug,normalized=True,trimmed=True,threshold=None)
for i in range(0,num_permutations):
try:
shuffled_data,shuffled_target = dfm.shuffle_scikit_data_target(scikit_data,scikit_target)
start_time = datetime.datetime.now()
cv.cross_val_score_filter_feature_selection(self.model,cv.trim_X_num_features,feature_size,shuffled_data,shuffled_target,cv=5)
end_time = datetime.datetime.now()
yield float((end_time - start_time).microseconds) / 100000
except:
yield 0.0
def get_predictions_full_CCLE_dataset_rfe(self,expression_file,ic50_file,target_features,drug):
scikit_data,scikit_target = dfm.get_expression_scikit_data_target_for_drug(expression_file,ic50_file,drug,normalized=True,trimmed=True,threshold=None)
step_length = int(len(scikit_data.tolist()[0]) / 100) + 1
model = RFE(self.model,target_features,step=step_length)
model.fit(scikit_data,scikit_target)
expression_frame = dfm.normalize_expression_frame(dfm.get_cell_line_expression_frame(expression_file))
cell_lines = expression_frame.columns
testing_data = dfm.get_scikit_data(expression_frame)
predictions = model.predict(testing_data)
top_features = [expression_frame.index[i] for i in xrange(0,len(expression_frame.index)) if model.support_[i]]
return cell_lines,predictions,top_features
