def train(ticker):
url = "https://www.alphavantage.co/query?function=TIME_SERIES_DAILY&symbol="
nextP = "&outputsize=full&apikey=T69K620H31T06293&datatype=csv"
r = requests.get(url + ticker + nextP)
path = 'temp/data/' + ticker + '.csv'
pathm = 'temp/models/' + ticker + '.h5'
with open(path, 'wb') as f:
f.write(r.content)
# FOR REPRODUCIBILITY
np.random.seed(7)
# IMPORTING DATASET
dataset = pd.read_csv(path, usecols=[1, 2, 3, 4])
dataset = dataset.reindex(index=dataset.index[::-1])
# CREATING OWN INDEX FOR FLEXIBILITY
obs = np.arange(1, len(dataset) + 1, 1)
# TAKING DIFFERENT INDICATORS FOR PREDICTION
OHLC_avg = dataset.mean(axis=1)
HLC_avg = dataset[['high', 'low', 'close']].mean(axis=1)
close_val = dataset[['close']]
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg), 1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(1, step_size), return_sequences=True))
model.add(LSTM(16))
model.add(Dense(1))
model.add(Activation('linear'))
# MODEL COMPILING AND TRAINING
# Try SGD, adam, adagrad and compare!!!
model.compile(loss='mean_squared_error', optimizer='adagrad')
model.fit(trainX, trainY, epochs=5, batch_size=1, verbose=2)
model.save(pathm)
return ticker + " Trained"
plt.plot(obs, close_val, 'g', label = 'Closing price')
plt.legend(loc = 'upper right')
plt.show()
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg),1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC,:], OHLC_avg[train_OHLC:len(OHLC_avg),:]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(1, step_size), return_sequences = True))
model.add(LSTM(16))
model.add(Dense(1))
model.add(Activation('linear'))
print(model.summary())
# Calculate average of Low and High values
OHLC_avg = np.reshape(OHLC_avg, (len(OHLC_avg), 2))
#Scale the values
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_datetime = int(len(datetime) * 0.75)
test_datetime = len(datetime) - train_datetime
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
train_datetime, test_datetime = datetime[
0:train_datetime, :], datetime[train_datetime:len(datetime), :]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 2)
testX, testY = preprocessing.new_dataset(test_OHLC, 2)
train_datetimeX, train_datetimeY = preprocessing.new_dataset(
train_datetime, 2)
test_datetimeX, test_datetimeY = preprocessing.new_dataset(
test_datetime, 2)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
train_datetime = np.reshape(
train_datetime,
(train_datetime.shape[0], 1, train_datetime.shape[1]))
test_datetime = np.reshape(
test_datetime, (test_datetime.shape[0], 1, test_datetime.shape[1]))
step_size = 2
def net(close_val, step):
nextValues = []
for i in range(step):
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(close_val.values, (len(close_val), 1))
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * .99)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(1, step_size), return_sequences=True))
model.add(LSTM(16))
model.add(Dense(1))
model.add(Activation('linear'))
# MODEL COMPILING AND TRAINING
model.compile(
loss='mean_squared_error',
optimizer='adagrad') # Try SGD, adam, adagrad and compare!!!
model.fit(trainX, trainY, epochs=5, batch_size=1, verbose=2)
# PREDICTION
trainPredict = model.predict(trainX)
testPredict = model.predict(testX)
# DE-NORMALIZING FOR PLOTTING
trainPredict = scaler.inverse_transform(trainPredict)
trainY = scaler.inverse_transform([trainY])
testPredict = scaler.inverse_transform(testPredict)
testY = scaler.inverse_transform([testY])
# TRAINING RMSE
trainScore = math.sqrt(
mean_squared_error(trainY[0], trainPredict[:, 0]))
print('Train RMSE: %.2f' % (trainScore))
# TEST RMSE
testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:, 0]))
print('Test RMSE: %.2f' % (testScore))
# CREATING SIMILAR DATASET TO PLOT TRAINING PREDICTIONS
trainPredictPlot = np.empty_like(OHLC_avg)
trainPredictPlot[:, :] = np.nan
trainPredictPlot[step_size:len(trainPredict) +
step_size, :] = trainPredict
OHLC_avg
# CREATING SIMILAR DATASSET TO PLOT TEST PREDICTIONS
testPredictPlot = np.empty_like(OHLC_avg)
testPredictPlot[:, :] = np.nan
testPredictPlot[len(trainPredict) + (step_size * 2) + 1:len(OHLC_avg) -
1, :] = testPredict
# DE-NORMALIZING MAIN DATASET
OHLC_avg = scaler.inverse_transform(OHLC_avg)
# PREDICT FUTURE VALUES
last_val = pd.Series(testPredict[-1])
close_val.append(last_val)
nextValues.append(last_val)
return nextValues
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg), 1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
step_size = 60
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, step_size)
testX, testY = preprocessing.new_dataset(test_OHLC, step_size)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], trainX.shape[1], 1))
testX = np.reshape(testX, (testX.shape[0], testX.shape[1], 1))
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(step_size, 1), return_sequences=True))
model.add(Dropout(0.02))
# model.add(LSTM(16, return_sequences=True))
# model.add(Dropout(0.02))
model.add(LSTM(16))
model.add(Dropout(0.02))
model.add(Dense(1))
def pred(ticker):
path = 'temp/data/' + ticker + '.csv'
pathm = 'temp/models/' + ticker + '.h5'
# IMPORTING DATASET
dataset = pd.read_csv(path, usecols=[1, 2, 3, 4])
dataset = dataset.reindex(index=dataset.index[::-1])
# CREATING OWN INDEX FOR FLEXIBILITY
obs = np.arange(1, len(dataset) + 1, 1)
# TAKING DIFFERENT INDICATORS FOR PREDICTION
OHLC_avg = dataset.mean(axis=1)
HLC_avg = dataset[['high', 'low', 'close']].mean(axis=1)
close_val = dataset[['close']]
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg), 1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
model = load_model(pathm)
# PREDICTION
trainPredict = model.predict(trainX)
testPredict = model.predict(testX)
# DE-NORMALIZING FOR PLOTTING
trainPredict = scaler.inverse_transform(trainPredict)
trainY = scaler.inverse_transform([trainY])
testPredict = scaler.inverse_transform(testPredict)
testY = scaler.inverse_transform([testY])
# TRAINING RMSE
trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:, 0]))
# print('Train RMSE: %.2f' % (trainScore))
# TEST RMSE
testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:, 0]))
# print('Test RMSE: %.2f' % (testScore))
# CREATING SIMILAR DATASET TO PLOT TRAINING PREDICTIONS
trainPredictPlot = np.empty_like(OHLC_avg)
trainPredictPlot[:, :] = np.nan
trainPredictPlot[step_size:len(trainPredict) + step_size, :] = trainPredict
# CREATING SIMILAR DATASSET TO PLOT TEST PREDICTIONS
testPredictPlot = np.empty_like(OHLC_avg)
testPredictPlot[:, :] = np.nan
testPredictPlot[len(trainPredict) + (step_size * 2) + 1:len(OHLC_avg) -
1, :] = testPredict
# DE-NORMALIZING MAIN DATASET
OHLC_avg = scaler.inverse_transform(OHLC_avg)
# PREDICT FUTURE VALUES
last_val = testPredict[-1]
last_val_scaled = last_val / last_val
next_val = model.predict(np.reshape(last_val_scaled, (1, 1, 1)))
finobj = {
'last': np.asscalar(last_val),
'next': np.asscalar(last_val * next_val)
}
r.set(ticker + "last", finobj['last'])
r.set(ticker + 'next', finobj['next'])
return {'result': ticker + " Values Stored"}
def lstm_neural_network(historical_data, look_back=240):
# FOR REPRODUCIBILITY
np.random.seed(7)
# IMPORTING DATASET
# dataset = pd.read_csv('apple_share_price.csv', usecols=[1,2,3,4])
dataset = historical_data.drop(['Date', 'Volume'], axis=1)
dataset = dataset.reindex(index=dataset.index[::-1])
# CREATING OWN INDEX FOR FLEXIBILITY
obs = np.arange(1, len(dataset) + 1, 1)
# TAKING DIFFERENT INDICATORS FOR PREDICTION
# OHLC_avg = dataset.mean(axis = 1)
HLC_avg = dataset[['High', 'Low', 'Close']].mean(axis=1)
close_val = dataset[['Close']]
OHLC_avg = dataset[['Close']]
# PLOTTING ALL INDICATORS IN ONE PLOT
# plt.plot(obs, OHLC_avg, 'r', label = 'OHLC avg')
# plt.plot(obs, HLC_avg, 'b', label = 'HLC avg')
# plt.plot(obs, close_val, 'g', label = 'Closing price')
# plt.legend(loc = 'upper right')
# plt.show()
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg), 1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, look_back)
testX, testY = preprocessing.new_dataset(test_OHLC, look_back)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
# # LSTM MODEL
# model = Sequential()
# model.add(LSTM(32, input_shape=(1, look_back), return_sequences = True))
# model.add(Dropout(0.1))
# model.add(LSTM(16))
# model.add(Dropout(0.1))
# model.add(Dense(1))
# model.add(Activation('linear'))
# # MODEL COMPILING AND TRAINING
# model.compile(loss='mean_squared_error', optimizer='adagrad') # Try SGD, adam, adagrad and compare!!!
# model.fit(trainX, trainY, epochs=1000, batch_size=1, verbose=2)
# LSTM MODEL 2
model = Sequential()
model.add(LSTM(25, input_shape=(1, look_back)))
model.add(Dropout(0.1))
model.add(Dense(1))
model.compile(loss='mse', optimizer='adam')
model.fit(trainX, trainY, epochs=1000, batch_size=240, verbose=2)
# PREDICTION
trainPredict = model.predict(trainX)
testPredict = model.predict(testX)
# DE-NORMALIZING FOR PLOTTING
trainPredict = scaler.inverse_transform(trainPredict)
trainY = scaler.inverse_transform([trainY])
testPredict = scaler.inverse_transform(testPredict)
testY = scaler.inverse_transform([testY])
# TRAINING RMSE
trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:, 0]))
print('Train RMSE: %.2f' % (trainScore))
# TEST RMSE
testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:, 0]))
print('Test RMSE: %.2f' % (testScore))
# CREATING SIMILAR DATASET TO PLOT TRAINING PREDICTIONS
trainPredictPlot = np.empty_like(OHLC_avg)
trainPredictPlot[:, :] = np.nan
trainPredictPlot[look_back:len(trainPredict) + look_back, :] = trainPredict
# CREATING SIMILAR DATASSET TO PLOT TEST PREDICTIONS
testPredictPlot = np.empty_like(OHLC_avg)
testPredictPlot[:, :] = np.nan
testPredictPlot[len(trainPredict) + (look_back * 2) + 1:len(OHLC_avg) -
1, :] = testPredict
# DE-NORMALIZING MAIN DATASET
OHLC_avg = scaler.inverse_transform(OHLC_avg)
# PLOT OF MAIN OHLC VALUES, TRAIN PREDICTIONS AND TEST PREDICTIONS
plt.plot(OHLC_avg, 'g', label='original dataset')
plt.plot(trainPredictPlot, 'r', label='training set')
plt.plot(testPredictPlot, 'b', label='predicted stock price/test set')
plt.legend(loc='upper right')
plt.xlabel('Time in Days')
plt.ylabel('OHLC Value of Apple Stocks')
# plt.show()
# PREDICT FUTURE VALUES
last_val = testPredict[-240:]
print(last_val.shape)
# last_val_scaled = last_val/np.linalg.norm(last_val)
last_val_scaled = scaler.fit_transform(last_val)
next_val = model.predict(np.reshape(last_val_scaled, (1, 1, 240)))
last_val_dnormalized = scaler.inverse_transform(last_val)
next_val_dnormalized = scaler.inverse_transform(next_val)
print("Last Candle Value: {}".format(np.asscalar(
last_val_dnormalized[-1])))
print("Next Candle value: {}".format(np.asscalar(
next_val_dnormalized[-1])))
plt.plot(np.asscalar(last_val_dnormalized[-1]),
'y',
label='next candle prediction')
plt.show()
return np.asscalar(last_val_dnormalized[-1]), np.asscalar(
next_val_dnormalized[-1])
def LSTMPredict(Stockname):
# FOR REPRODUCIBILITY
np.random.seed(7)
engine = create_engine(
'mysql+mysqlconnector://root:123456@localhost:3306/TESTDB')
df = pd.read_sql_query('SELECT * FROM ' + Stockname + '_History_Price',
engine)
#dataset = pd.read_csv('apple_share_price.csv', usecols=[1,2,3,4])
# IMPORTING DATASET
#dataset=df[['R_Price']]
dataset = df[['Open_Price', 'High_Price', 'Low_Price', 'Close_Price']]
dataset = dataset.apply(pd.to_numeric)
dataset = dataset.reindex(index=dataset.index[::-1])
# CREATING OWN INDEX FOR FLEXIBILITY
obs = np.arange(1, len(dataset) + 1, 1)
# TAKING DIFFERENT INDICATORS FOR PREDICTION
OHLC_avg = dataset.mean(axis=1)
HLC_avg = dataset[['High_Price', 'Low_Price', 'Close_Price']].mean(axis=1)
close_val = dataset[['Close_Price']]
# PLOTTING ALL INDICATORS IN ONE PLOT
plt.plot(obs, OHLC_avg, 'r', label='OHLC avg')
plt.plot(obs, HLC_avg, 'b', label='HLC avg')
plt.plot(obs, close_val, 'g', label='Closing price')
plt.legend(loc='upper right')
plt.show()
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg), 1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC, :], OHLC_avg[
train_OHLC:len(OHLC_avg), :]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(1, step_size), return_sequences=True))
model.add(LSTM(16))
model.add(Dense(1))
model.add(Activation('linear'))
# MODEL COMPILING AND TRAINING
model.compile(loss='mean_squared_error',
optimizer='adagrad') # Try SGD, adam, adagrad and compare!!!
model.fit(trainX, trainY, epochs=100, batch_size=1, verbose=2)
# PREDICTION
trainPredict = model.predict(trainX)
testPredict = model.predict(testX)
# DE-NORMALIZING FOR PLOTTING
trainPredict = scaler.inverse_transform(trainPredict)
trainY = scaler.inverse_transform([trainY])
testPredict = scaler.inverse_transform(testPredict)
testY = scaler.inverse_transform([testY])
# TRAINING RMSE
trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:, 0]))
print('Train RMSE: %.2f' % (trainScore))
# TEST RMSE
testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:, 0]))
print('Test RMSE: %.2f' % (testScore))
print(mean_absolute_error(testY[0], testPredict[:, 0]))
# CREATING SIMILAR DATASET TO PLOT TRAINING PREDICTIONS
trainPredictPlot = np.empty_like(OHLC_avg)
trainPredictPlot[:, :] = np.nan
trainPredictPlot[step_size:len(trainPredict) + step_size, :] = trainPredict
# CREATING SIMILAR DATASSET TO PLOT TEST PREDICTIONS
testPredictPlot = np.empty_like(OHLC_avg)
testPredictPlot[:, :] = np.nan
testPredictPlot[len(trainPredict) + (step_size * 2) + 1:len(OHLC_avg) -
1, :] = testPredict
avg = testPredict.mean(axis=1)
av = avg.mean(axis=0)
# DE-NORMALIZING MAIN DATASET
OHLC_avg = scaler.inverse_transform(OHLC_avg)
# PLOT OF MAIN OHLC VALUES, TRAIN PREDICTIONS AND TEST PREDICTIONS
plt.plot(OHLC_avg, 'g', label='original dataset')
plt.plot(trainPredictPlot, 'r', label='training set')
plt.plot(testPredictPlot, 'b', label='predicted stock price/test set')
plt.legend(loc='upper right')
plt.xlabel('Time in Days')
plt.ylabel('OHLC Value of Stocks')
plt.show()
# PREDICT FUTURE VALUES
last_val = testPredict[-1]
last_val_scaled = last_val / last_val
next_val = model.predict(np.reshape(last_val_scaled, (1, 1, 1)))
print("Last Day Value:", np.asscalar(last_val))
print("Next Day Value:", np.asscalar(last_val * next_val))
def StockPricePrediction():
# IMPORTING IMPORTANT LIBRARIES
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import math
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import mean_squared_error
from keras.models import Sequential
from keras.models import load_model
from keras.layers import Dense, Activation
from keras.layers import LSTM
import preprocessing
import io
import base64
# FOR REPRODUCIBILITY
np.random.seed(7) # This is done to prevent the use of randomness, such as initializing to random weights, and in turn the same network trained on the same data can produce different results
# IMPORTING DATASET
dataset = pd.read_csv('IDBI.csv', usecols=[1,2,3,4])
dataset = dataset.reindex(index = dataset.index[::-1]) #transpose the matrix and fills with boolean values
# CREATING OWN INDEX FOR FLEXIBILITY
obs = np.arange(1, len(dataset) + 1, 1) #builds an array having content from 1 to 1664(Number of rows in dataset)
# TAKING DIFFERENT INDICATORS FOR PREDICTION
OHLC_avg = dataset.mean(axis = 1) # each row average is taken and stored in respective index
HLC_avg = dataset[['High', 'Low', 'Close']].mean(axis = 1)
close_val = dataset[['Close']]
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg),1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.75) # Train is 75 % and test is 25 %
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC,:], OHLC_avg[train_OHLC:len(OHLC_avg),:]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(1, step_size), return_sequences = True))
model.add(LSTM(16))
model.add(Dense(1))
model.add(Activation('linear'))
# MODEL COMPILING AND TRAINING
model.compile(loss='mean_squared_error', optimizer='adagrad') # Try SGD, adam, adagrad and compare!!!
model.fit(trainX, trainY, epochs=5, batch_size=1, verbose=2)
# PREDICTION
trainPredict = model.predict(trainX)
testPredict = model.predict(testX)
# DE-NORMALIZING FOR PLOTTING
trainPredict = scaler.inverse_transform(trainPredict)
trainY = scaler.inverse_transform([trainY])
testPredict = scaler.inverse_transform(testPredict)
testY = scaler.inverse_transform([testY])
# TRAINING RMSE
trainScore = math.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))
print('Train RMSE: %.2f' % (trainScore))
# TEST RMSE
testScore = math.sqrt(mean_squared_error(testY[0], testPredict[:,0]))
print('Test RMSE: %.2f' % (testScore))
# CREATING SIMILAR DATASET TO PLOT TRAINING PREDICTIONS
trainPredictPlot = np.empty_like(OHLC_avg)
trainPredictPlot[:, :] = np.nan
trainPredictPlot[step_size:len(trainPredict)+step_size, :] = trainPredict
# CREATING SIMILAR DATASSET TO PLOT TEST PREDICTIONS
testPredictPlot = np.empty_like(OHLC_avg)
testPredictPlot[:, :] = np.nan
testPredictPlot[len(trainPredict)+(step_size*2)+1:len(OHLC_avg)-1, :] = testPredict
# DE-NORMALIZING MAIN DATASET
OHLC_avg = scaler.inverse_transform(OHLC_avg)
img = io.BytesIO()
# PLOTTING ALL INDICATORS IN ONE PLOT
plt.plot(obs, OHLC_avg, 'r', label = 'OHLC avg')
plt.plot(obs, HLC_avg, 'b', label = 'HLC avg')
plt.plot(obs, close_val, 'g', label = 'Closing price')
plt.legend(loc = 'upper left')
plt.xlabel('Time in Days')
plt.ylabel('OHLC Value of Apple Stocks')
plt.savefig(img, format='png')
img.seek(0)
plot_url1 = base64.b64encode(img.getvalue()).decode()
# PLOT OF MAIN OHLC VALUES, TRAIN PREDICTIONS AND TEST PREDICTIONS
plt.plot(OHLC_avg, 'g', label = 'original dataset')
plt.plot(trainPredictPlot, 'r', label = 'training set')
plt.plot(testPredictPlot, 'b', label = 'predicted stock price/test set')
plt.legend(loc = 'upper left')
plt.xlabel('Time in Days')
plt.ylabel('OHLC Value of Apple Stocks')
plt.legend(loc = 'upper left')
plt.savefig(img, format='png')
img.seek(0)
plot_url2 = base64.b64encode(img.getvalue()).decode()
return '<h1>IDBI Bank - Actual Stock Price</h1><img src="data:image/png;base64,{}"><hl><h1>IDBI Bank Predicted Stock Price</h1><img src="data:image/png;base64,{}">'.format(plot_url1, plot_url2)
# PREDICT FUTURE VALUES
last_val = testPredict[-1]
last_val_scaled = last_val/last_val
next_val = model.predict(np.reshape(last_val_scaled, (1,1,1)))
print ("Last Day Value:", np.asscalar(last_val))
print ("Next Day Value:", np.asscalar(last_val*next_val))
# plt.plot(close_val, 'g', label = 'Closing price')
# plt.legend(loc = 'upper right')
# plt.show()
# PREPARATION OF TIME SERIES DATASE
OHLC_avg = np.reshape(OHLC_avg.values, (len(OHLC_avg),1)) # 1664
scaler = MinMaxScaler(feature_range=(0, 1))
OHLC_avg = scaler.fit_transform(OHLC_avg)
# TRAIN-TEST SPLIT
train_OHLC = int(len(OHLC_avg) * 0.70)
test_OHLC = len(OHLC_avg) - train_OHLC
train_OHLC, test_OHLC = OHLC_avg[0:train_OHLC,:], OHLC_avg[train_OHLC:len(OHLC_avg),:]
# TIME-SERIES DATASET (FOR TIME T, VALUES FOR TIME T+1)
trainX, trainY = preprocessing.new_dataset(train_OHLC, 1)
testX, testY = preprocessing.new_dataset(test_OHLC, 1)
# RESHAPING TRAIN AND TEST DATA
trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
step_size = 1
# LSTM MODEL
model = Sequential()
model.add(LSTM(32, input_shape=(1, step_size), return_sequences = True))
model.add(LSTM(16))
model.add(Dense(1))
model.add(Activation('linear'))
# MODEL COMPILING AND TRAINING
