import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential, load_model
from keras.layers import LSTM, Dense, Dropout
import os

"""### Loading in Dataset"""

df = pd.read_csv('/content/Data.csv')
df.head()

"""### Preprocessing and Feature Extraction"""

df = df['Close'].values
df = df.reshape(-1, 1)
print(df.shape)
df[:5]

dataset_train = np.array(df[:int(df.shape[0]*0.8)])
dataset_test = np.array(df[int(df.shape[0]*0.8)-100:])
print(dataset_train.shape)
print(dataset_test.shape)

scaler = MinMaxScaler(feature_range=(0,1))
dataset_train = scaler.fit_transform(dataset_train)
dataset_train[:5]

dataset_test = scaler.transform(dataset_test)
dataset_test[:5]

def create_dataset(df):
    x = []
    y = []
    for i in range(100, df.shape[0]):
        x.append(df[i-100:i, 0])
        y.append(df[i, 0])
    x = np.array(x)
    y = np.array(y)
    return x,y

x_train, y_train = create_dataset(dataset_train)
x_train[:1]

x_test, y_test = create_dataset(dataset_test)
x_test[:1]

# Reshape features for LSTM Layer
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))

"""### Building Model"""

model = Sequential()
model.add(LSTM(units=96, return_sequences=True, input_shape=(x_train.shape[1], 1)))
model.add(Dropout(0.2))
model.add(LSTM(units=96, return_sequences=True))
model.add(Dropout(0.2))
model.add(LSTM(units=96, return_sequences=True))
model.add(Dropout(0.2))
model.add(LSTM(units=96))
model.add(Dropout(0.2))
model.add(Dense(units=1))

model.compile(loss='mean_squared_error', optimizer='adam')

if(not os.path.exists('stock_prediction.h5')):
    model.fit(x_train, y_train, epochs=1, batch_size=32)
    model.save('stock_prediction.h5')

from google.colab import drive
drive.mount('/content/drive')

model = load_model('stock_prediction.h5')

"""### Visualizing Results"""

predictions = model.predict(x_test)
predictions = scaler.inverse_transform(predictions)

fig, ax = plt.subplots(figsize=(8,4))
plt.plot(df, color='red',  label="True Price")
ax.plot(range(len(y_train)+100,len(y_train)+100+len(predictions)),predictions, color='blue', label='Predicted Testing Price')
plt.legend()

y_test_scaled = scaler.inverse_transform(y_test.reshape(-1, 1))

model = Sequential()
model.add(LSTM(x_train.shape[1], return_sequences=True, input_shape=[x_train.shape[1], 1]))
model.add(LSTM(x_train.shape[1], return_sequences=False))
model.add(Dense(1))
model.compile(optimizer='adam', loss = 'mean_squared_error')
model.fit(x_train, y_train, batch_size = 1, epochs=30)


fig, ax = plt.subplots(figsize=(8,4))
ax.plot(y_test_scaled, color='red', label='True Testing Price')
plt.plot(predictions, color='blue', label='Predicted Testing Price')
plt.legend()

x = x_test[-1]
num_timesteps = 1
preds = []
for i in range(num_timesteps):
    data = np.expand_dims(x, axis=0)
    prediction = model.predict(data)
    prediction = scaler.inverse_transform(prediction)
    preds.append(prediction[0][0])
    x = np.delete(x, 0, axis=0) # delete first row
    x = np.vstack([x, prediction]) # add prediction

print(preds)