T = basic.get_X_from_dataframe(test_set)
Id = basic.get_Id_from_test_dataframe(test_set)
#Get a classifier. Later, it will be a pipeline
pipeline = bayes.GaussianNB()
#TODO tune pipeline metaparameters.
#The most time consuming part! We have to be really creative here!
#Once the metaparameters were found, set them and fit ("learn") the training set
pipeline.fit(X,Y)
#Sometimes a little time consuming but not very much
predicted_proba = pipeline.predict_proba(T)
#Build nicely the pandas so we can write a nice csv
output = basic.build_pandas_output(Id,predicted_proba,class_labels)
output_file_name = fh.get_out_path("cv_Prediction_Attempt_01.08.2015.csv")
output.to_csv(output_file_name,index=False)
#Write the file, ready to be submitted
print "Done"
def test_how_to_write_a_file():
file_name = fh.get_out_path("sample_writable_file.txt")
with open(file_name,"w") as file:
file.write("Some string in the file\n")
