prepare_data.py

training_data = "data/training/training-data.csv"
training_label = "data/training/training-label.csv"


X_dataframe = pandas.read_csv(training_data, header=None)
X = X_dataframe.values
Y_dataframe = pandas.read_csv(training_label, header=None)
Y = Y_dataframe.values

dummy_x = []
for text in X:
	dummy_x.append(numpy.array(tobow(text[0])[0]))

bow = numpy.array(dummy_x)

# encode class values as integers
encoder = LabelEncoder()
encoder.fit(Y)
numpy.save(encoder_path,encoder.classes_)
encoded_Y = encoder.transform(Y)
# convert integers to dummy variables (i.e. one hot encoded)
dummy_y = np_utils.to_categorical(encoded_Y)