import keras
from keras import backend as K
from keras.models import Sequential
from keras.layers import Activation
from keras.layers import LSTM
from keras.layers.core import Dense
from keras.optimizers import Adam
from keras.metrics import categorical_crossentropy

import numpy as np
from keras.models import Sequential
from sklearn.preprocessing import MinMaxScaler



ACTIONS_DIM = 2
OBSERVATIONS_DIM = 4
MAX_ITERATIONS = 200
LEARNING_RATE = 0.001

def data1 ():
    data = np.genfromtxt("2.txt", delimiter=",")
    print data
    return data

def samples1(i):
    for b in range (5):
        print "A"
        if (data [b] == ","):
            print "A"
            return 0
        print "B"
        return 0

def get_model():
    model = Sequential([
        Dense(16, input_shape=(1,), activation="relu"),
        Dense(32, activation="relu"),
        Dense(2, activation="softmax"),
    ])

    model.compile(
        optimizer=Adam(lr=LEARNING_RATE),
        loss='sparse_categorical_crossentropy',
        metrics=["accuracy"],
    )

    model.summary()

    return model

def get_model2():

    model = Sequential()
    model.add(LSTM(64, input_shape=(1,64), return_sequences=True))
    model.add(Dense(64))
    model.compile(loss='mean_absolute_error', optimizer='adam',metrics=['accuracy'])
    model.summary()

    return model

def train2(model, scaled_train_samples2, train_labels):
    model.fit(scaled_train_samples2, train_labels, nb_epoch=100, batch_size=1, verbose=2)


def train(model, scaled_train_samples, train_labels):
    model.fit(scaled_train_samples, train_labels, batch_size = 100, epochs=10, verbose=2)

def predict(model, train_samples):
  return model.predict(train_samples)

def main():


    SAMPLES = 65
    data = []
    samples = []
    train_samples = []
    train_labels = []
    predict_samples = []
    predict_labels = []


    data = data1()

    Date    =data[:,][:,0]
    Open    =data[:,][:,1]
    High    =data[:,][:,2]
    Low     =data[:,][:,3]
    Close   =data[:,][:,4]
    Volume  =data[:,][:,5]
    OpenInt =data[:,][:,6]


    print   Date
    print   Open
    print   High
    print   Low
    print   Close
    print   Volume
    print   OpenInt

    for i in range(SAMPLES):
#       samples.append(Open[i])
        samples.append(High[i])
        samples.append(Low[i])
#       samples.append(Close[i])

    print samples

    for i in range(SAMPLES-1):
        if (samples[i+1] >= samples[i]):
            a=1
        else:
            a=0
        train_samples.append(samples[i])
        train_labels.append(a)

    train_samples = np.array(train_samples)
    train_labels = np.array(train_labels)
    train_labels2 = train_labels.reshape((1,1,SAMPLES-1))

    scaler = MinMaxScaler(feature_range=(0,1))

    scaled_train_samples = scaler.fit_transform((train_samples).reshape(-1,1))
    scaled_train_samples2 = train_samples.reshape((1,1,SAMPLES-1))

    print scaled_train_samples
    print train_labels

    model = get_model()
    model2 = get_model2()

    train(model, scaled_train_samples, train_labels)
    print scaled_train_samples2
    train2(model2, scaled_train_samples2, train_labels2)

    for i in range(100):
        if (samples[i+SAMPLES+1] >= samples[i+SAMPLES]):
            a=1
        else:
            a=0
        predict_samples.append(samples[i])
        predict_labels.append(a)

    predict_samples = np.array(predict_samples)
    predict_labels = np.array(predict_labels)

    scaler = MinMaxScaler(feature_range=(0,1))

    scaled_predict_samples = scaler.fit_transform((predict_samples).reshape(-1,1))

    print predict(model,scaled_predict_samples)

    print "hallo"
    return 0

if __name__ == '__main__':
    main()