sinkeras.py

import os
import time
import warnings
import math
import numpy as np
from numpy import newaxis
from keras.layers.core import Dense, Activation, Dropout
from keras.layers.recurrent import LSTM
from keras.models import Sequential
import matplotlib.pyplot as plt

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3' #Hide messy TensorFlow warnings
warnings.filterwarnings("ignore") #Hide messy Numpy warnings

epochs  = 1
seq_len = 50

def load_data(seq_len):        
    data = [math.sin(x/5)+1 for x in range(1000)]
    
    sequence_length = seq_len + 1
    result = []
    for index in range(len(data) - sequence_length):
        result.append(data[index: index + sequence_length])

    result = np.array(result)

    row = round(0.9 * result.shape[0])    
    train = result[:int(row), :]
    np.random.shuffle(train)
    x_train = train[:, :-1]
#     y_train = train[:, -1]
    y_train = np.full((row), 1)
    x_test = result[int(row):, :-1]
    y_test = result[int(row):, -1]   

    x_train = np.reshape(x_train, (x_train.shape[0], x_train.shape[1], 1))
    x_test = np.reshape(x_test, (x_test.shape[0], x_test.shape[1], 1))  

    return [x_train, y_train, x_test, y_test]

  
def build_model(layers):
    model = Sequential()

    model.add(LSTM(
        input_shape=(layers[1], layers[0]),
        output_dim=layers[1],
        return_sequences=False))
    model.add(Dropout(0.2))

    model.add(Dense(
        output_dim=layers[2]))
    model.add(Activation("linear"))

    model.compile(loss="mse", optimizer="rmsprop")
    return model


X_train, y_train, X_test, y_test = load_data(seq_len)

model = build_model([1, 50, 1])
model.fit(
    X_train,
    y_train,
    batch_size=512,
    nb_epoch=5,
    validation_split=0.05)
  

# plt.plot(model.predict(X_test))
# plt.show()

print(model.predict(X_test))