ML_file (1).py.py

# -*- coding: utf-8 -*-
"""FYP_iteration_3_stcked_lstm_tingo.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1cY9h_7j7V6FoVIVdfhzaLlByvjs82bS7

### Stock Market Prediction And Forecasting Using Stacked LSTM
"""

# !pip install tf-nightly
import matplotlib.pyplot as plt
import pandas as pd
import pandas_datareader as pdr

import numpy as np
import tensorflow as tf
tf.__version__
# import matplotlib.pyplot as plt
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import LSTM



# stock = 'RELIANCE.NS'  # 1
# stock = 'HDFCBANK.NS'  # 2
# stock = 'M&M.NS'  # 3
# stock = 'TATAMOTORS.NS'  # 4
# stock = 'BAJFINANCE.NS'  # 5
# stock = 'INFY'  # 6
# stock = 'ICICIBANK.NS'  # 7
# stock = 'BHEL.NS'  # 8
stock = 'ADANIPOWER.NS'  # 9
# stock = '(TCS.NS')  # 10


# stock_data = stock_data.loc['2018-5-31':]
# df = stock_data
# df.to_csv('/content/{}_dataset.csv'.format(stock), index=False)
stock_data = pd.read_csv('/content/{}_dataset.csv'.format(stock)) # put adress where csv files are stored
stock_data

# stock_data

stock_data=df.reset_index()['Close']

# stock_data #important


# plt.plot(stock_data) #imp

from sklearn.preprocessing import MinMaxScaler
scaler=MinMaxScaler(feature_range=(0,1))
stock_data=scaler.fit_transform(np.array(stock_data).reshape(-1,1))

# print(df1)

##splitting dataset into train and test split
training_size=int(len(stock_data)*0.65)
test_size=len(stock_data)-training_size
train_data,test_data=stock_data[0:training_size,:],stock_data[training_size:len(stock_data),:1]

# training_size,test_size #important

#train_data

import numpy
# convert an array of values into a dataset matrix
def create_dataset(dataset, time_step=1):
	dataX, dataY = [], []
	for i in range(len(dataset)-time_step-1):
		a = dataset[i:(i+time_step), 0]   ###i=0, 0,1,2,3-----99   100 
		dataX.append(a)
		dataY.append(dataset[i + time_step, 0])
	return numpy.array(dataX), numpy.array(dataY)

# reshape into X=t,t+1,t+2,t+3 and Y=t+4
time_step = 100
X_train, y_train = create_dataset(train_data, time_step)
X_test, ytest = create_dataset(test_data, time_step)

# print(X_train.shape), print(y_train.shape)#important

# print(X_test.shape), print(ytest.shape) #important

# reshape input to be [samples, time steps, features] which is required for LSTM
X_train =X_train.reshape(X_train.shape[0],X_train.shape[1] , 1)
X_test = X_test.reshape(X_test.shape[0],X_test.shape[1] , 1)

### Create the Stacked LSTM model

model=Sequential()
model.add(LSTM(50,return_sequences=True,input_shape=(time_step,1)))
model.add(LSTM(50,return_sequences=True))
model.add(LSTM(50))
model.add(Dense(1))
model.compile(loss='mean_squared_error',optimizer='adam')

model.summary()

model.fit(X_train,y_train,validation_data=(X_test,ytest),epochs=10,batch_size=64,verbose=1)

### Lets Do the prediction and check performance metrics
train_predict=model.predict(X_train)
test_predict=model.predict(X_test)

##Transformback to original form
train_predict=scaler.inverse_transform(train_predict)
test_predict=scaler.inverse_transform(test_predict)

### Calculate RMSE performance metrics
import math
from sklearn.metrics import mean_squared_error
math.sqrt(mean_squared_error(y_train,train_predict))

### Test Data RMSE
math.sqrt(mean_squared_error(ytest,test_predict))

### Plotting 
# shift train predictions for plotting
look_back=time_step
trainPredictPlot = numpy.empty_like(stock_data)
trainPredictPlot[:, :] = np.nan
trainPredictPlot[look_back:len(train_predict)+look_back, :] = train_predict
# shift test predictions for plotting
testPredictPlot = numpy.empty_like(stock_data)
testPredictPlot[:, :] = numpy.nan
testPredictPlot[len(train_predict)+(look_back*2)+1:len(stock_data)-1, :] = test_predict
# plot baseline and predictions
plt.plot(scaler.inverse_transform(stock_data))
plt.plot(trainPredictPlot)
plt.plot(testPredictPlot)
plt.show()




l = len(test_data) #important

x_input=test_data[l-100:].reshape(1,-1)
x_input.shape

temp_input=list(x_input)
temp_input=temp_input[0].tolist()

#temp_input
days = 30
n_steps=100

# demonstrate prediction for next 10 days
from numpy import array

lst_output=[]
i=0
while(i<days):
    
    if(len(temp_input)>100):
        #print(temp_input)
        x_input=np.array(temp_input[1:])
        # print("{} day input {}".format(i,x_input))#important
        x_input=x_input.reshape(1,-1)
        x_input = x_input.reshape((1, n_steps, 1))
        #print(x_input)
        yhat = model.predict(x_input, verbose=0)
        # print("{} day output {}".format(i,yhat))#important
        temp_input.extend(yhat[0].tolist())
        temp_input=temp_input[1:]
        #print(temp_input)
        lst_output.extend(yhat.tolist())
        i=i+1
    else:
        x_input = x_input.reshape((1, n_steps,1))
        yhat = model.predict(x_input, verbose=0)
        # print(yhat[0])#important
        temp_input.extend(yhat[0].tolist())
        # print(len(temp_input))#important
        lst_output.extend(yhat.tolist())
        i=i+1
    

# print(lst_output)

day_new=np.arange(1,time_step +1)
day_pred=np.arange(time_step +1,time_step +1 + days)

l2 = len(stock_data)

plt.plot(day_new,scaler.inverse_transform(stock_data[l2-100:]))
plt.plot(day_pred,scaler.inverse_transform(lst_output))

stock_data_merged=stock_data.tolist()
stock_data_merged.extend(lst_output)
plt.plot(stock_data_merged[1200:])

stock_data_merged=scaler.inverse_transform(stock_data_merged).tolist()

plt.plot(stock_data_merged)

from datetime import datetime, timedelta

def generate_dates(start_date, n):
    dates = []
    current_date = start_date
    
    for _ in range(n):
        dates.append(current_date.strftime('%Y-%m-%d'))
        current_date += timedelta(days=1)
    
    return dates

# Example usage
start_date = datetime(2023, 5, 31)  # Specify the start date
n = 10  # Specify the number of days

pred_date_list = generate_dates(start_date, n)
print(pred_date_list)

# original_dates = df['date'].to_list()
# original_dates

predicted_values = scaler.inverse_transform(lst_output)
# predicted_values

model.save('models/{}_stock-100_epochs.pb'.format(stock))

# #tflite
# model_for_lite = tf.keras.models.load_model('models/{}_stock-100_epochs.pb'.format(stock))
# # converter = tf.lite.TFLiteConverter.from_keras_model(model)

# converter._experimental_lower_tensor_list_ops = False
# converter = tf.lite.TFLiteConverter.from_keras_model(model_for_lite)

# tflite_model = converter.convert()
# # open("models/APPl_stock-100_epochs.tflite","wb").write(tflite_model)

# with open("models/APPl_stock-100_epochs.tflite","wb") as f:
#     f.write(tflite_model)

# run_model = tf.function(lambda x: model(x))
# # This is important, let's fix the input size.
# BATCH_SIZE = 1
# STEPS = 100
# INPUT_SIZE = 50
# concrete_func = run_model.get_concrete_function(
#     tf.TensorSpec([BATCH_SIZE, STEPS, INPUT_SIZE], model.inputs[0].dtype))

# # model directory.
# MODEL_DIR = "keras_lstm"
# model.save(MODEL_DIR, save_format="tf", signatures=concrete_func)

# converter = tf.lite.TFLiteConverter.from_saved_model(MODEL_DIR)
# tflite_model = converter.convert()

close = df['Close']
open = df['Open']
high = df['High']
low = df['Low']
# close = close.to_list()
# type(close)
close
open
high
low
predictedpoints = scaler.inverse_transform(lst_output)