import numpy
import pandas
from keras.models import Sequential
from keras.layers import Dense, Dropout
from keras.wrappers.scikit_learn import KerasClassifier
from keras.utils import np_utils
from sklearn.preprocessing import LabelEncoder
from tobow import tobow

numpy.random.seed(7)

model_path = "classifier/classifier.json"
weights_path = "classifier/weights.h5"
encoder_path = "classifier/encoder.npy"

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)

# define baseline model
def baseline_model():
	# create model
	model = Sequential()
	model.add(Dense(5000, input_shape=(len(bow[0]),), activation='relu'))
	model.add(Dropout(0.2))
	model.add(Dense(len(dummy_y[0]), activation='softmax'))
	# Compile model
	model.summary()
	model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
	return model

model = baseline_model()
#training process
model.fit(bow, dummy_y, epochs=30)

#Save to Json
model_json = model.to_json()
with open(model_path, "w") as json_file:
    json_file.write(model_json)
model.save_weights(weights_path)
print "Model has been saved"