!pip install tensorflow-gpu
!pip install yfinance --upgrade --no-cache-dir
!pip install pandas_datareader
!pip install sklearn
!pip install -U protobuf==3.8.0import math
import pandas_datareader.data as web
from datetime import date, timedelta
import pandas as pd
from numpy import array, reshape
from keras.models import Sequential
from keras.layers import LSTM, SimpleRNN, Dropout, Dense
import matplotlib.pyplot as plt
from sklearn.metrics import accuracy_score
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import mean_absolute_error
from sklearn.model_selection import train_test_split
from keras.callbacks import EarlyStopping
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras.layers import Dropout
import yfinance as yf
yf.pdr_override()b3sa3 = web.get_data_yahoo("B3SA3.SA",start="2000-01-01", end="2022-12-31")plt.figure(figsize = (12,6))
plt.plot(b3sa3["Open"])
plt.plot(b3sa3["High"])
plt.plot(b3sa3["Low"])
plt.plot(b3sa3["Close"])
plt.title('Tesla stock price history')
plt.ylabel('Price (USD)')
plt.xlabel('Days')
plt.legend(['Open','High','Low','Close'], loc='upper left')
plt.show()plt.figure(figsize = (12,6))
plt.plot(b3sa3["Open"])
plt.plot(b3sa3["High"])
plt.plot(b3sa3["Low"])
plt.plot(b3sa3["Close"])
plt.title('Tesla stock price history')
plt.ylabel('Price (USD)')
plt.xlabel('Days')
plt.legend(['Open','High','Low','Close'], loc='upper left')
plt.show()
plt.figure(figsize = (12,6))
plt.plot(b3sa3["Volume"])
plt.title('Tesla stock volume history')
plt.ylabel('Volume')
plt.xlabel('Days')
plt.show()
data_target = b3sa3.filter(['Close'])
target = data_target.values
training_data_len = math.ceil(len(target)* 0.75)
training_data_len
sc = MinMaxScaler(feature_range=(0,1))
training_scaled_data = sc.fit_transform(target)
training_scaled_data# split a univariate sequence into samples
def split_sequence(sequence, n_steps, avanco):
X, y = list(), list()
for i in range(len(sequence)):
# find the end of this pattern
end_ix = i + n_steps
z=end_ix-1
# check if we are beyond the sequence
if end_ix+avanco > len(sequence)-1:
break
# gather input and output parts of the pattern
seq_x, seq_y = sequence[i:end_ix], sequence[end_ix+avanco]
e=seq_y
f=seq_x[-1]
faixa = ((e/f)-1)*100
if faixa <-15:
faixa=-3
elif faixa <-5:
faixa=-2
elif faixa <0:
faixa=-1
elif faixa <5:
faixa=1
elif faixa <15:
faixa=2
elif faixa >15:
faixa=3
X.append(seq_x)
print(faixa)
y.append(faixa)
return array(X), array(y)train_data = training_scaled_data[0:training_data_len , : ]
X_train = []
y_train = []
for i in range(180, len(train_data)):
X_train.append(train_data[i-180:i, 0])
y_train.append(train_data[i, 0])
print(X_train)
X_train, y_train = array(X_train), array(y_train)
X_train = reshape(X_train, (X_train.shape[0], X_train.shape[1], 1))
print('Number of rows and columns: ', X_train.shape)model = Sequential()
#Adding the first LSTM layer and some Dropout regularisation
model.add(LSTM(units = 50, return_sequences = True, input_shape = (X_train.shape[1], 1)))
model.add(Dropout(0.2))
# Adding a second LSTM layer and some Dropout regularisation
model.add(LSTM(units = 50, return_sequences = True))
model.add(Dropout(0.2))
# Adding a third LSTM layer and some Dropout regularisation
model.add(LSTM(units = 50, return_sequences = True))
model.add(Dropout(0.2))
# Adding a fourth LSTM layer and some Dropout regularisation
model.add(LSTM(units = 50))
model.add(Dropout(0.2))
# Adding the output layer
model.add(Dense(units = 1))
# Compiling
model.compile(optimizer='adam', loss='mean_squared_error', metrics=['mae', 'acc'])
# Fitting the RNN to the Training set
model.fit(X_train, y_train, epochs = 100, batch_size = 32)Epoch 1/100 83/83 [==============================] - 19s 164ms/step - loss: 0.0022 - mae: 0.0293 - acc: 3.7864e-04 Epoch 2/100 83/83 [==============================] - 14s 171ms/step - loss: 7.5855e-04 - mae: 0.0168 - acc: 3.7864e-04 Epoch 3/100 83/83 [==============================] - 15s 183ms/step - loss: 6.4637e-04 - mae: 0.0156 - acc: 3.7864e-04 Epoch 4/100 83/83 [==============================] - 15s 176ms/step - loss: 6.5974e-04 - mae: 0.0156 - acc: 3.7864e-04 Epoch 5/100 83/83 [==============================] - 15s 178ms/step - loss: 5.6503e-04 - mae: 0.0149 - acc: 3.7864e-04 Epoch 6/100 83/83 [==============================] - 15s 179ms/step - loss: 5.4539e-04 - mae: 0.0154 - acc: 3.7864e-04 Epoch 7/100 83/83 [==============================] - 15s 179ms/step - loss: 4.2728e-04 - mae: 0.0131 - acc: 3.7864e-04 Epoch 8/100 83/83 [==============================] - 15s 180ms/step - loss: 4.4834e-04 - mae: 0.0133 - acc: 3.7864e-04 Epoch 9/100 83/83 [==============================] - 15s 184ms/step - loss: 4.4234e-04 - mae: 0.0130 - acc: 3.7864e-04 Epoch 10/100 83/83 [==============================] - 15s 186ms/step - loss: 4.5409e-04 - mae: 0.0139 - acc: 3.7864e-04 Epoch 11/100 83/83 [==============================] - 16s 197ms/step - loss: 4.1856e-04 - mae: 0.0127 - acc: 3.7864e-04 Epoch 12/100 83/83 [==============================] - 16s 187ms/step - loss: 4.0557e-04 - mae: 0.0131 - acc: 3.7864e-04 Epoch 13/100 83/83 [==============================] - 15s 185ms/step - loss: 3.6888e-04 - mae: 0.0120 - acc: 3.7864e-04 Epoch 14/100 83/83 [==============================] - 15s 186ms/step - loss: 3.8366e-04 - mae: 0.0127 - acc: 3.7864e-04 Epoch 15/100 83/83 [==============================] - 16s 189ms/step - loss: 3.4191e-04 - mae: 0.0118 - acc: 3.7864e-04 Epoch 16/100 83/83 [==============================] - 15s 178ms/step - loss: 3.4022e-04 - mae: 0.0118 - acc: 3.7864e-04 Epoch 17/100 83/83 [==============================] - 15s 175ms/step - loss: 3.4134e-04 - mae: 0.0122 - acc: 3.7864e-04 Epoch 18/100 83/83 [==============================] - 15s 178ms/step - loss: 3.3549e-04 - mae: 0.0118 - acc: 3.7864e-04 Epoch 19/100 83/83 [==============================] - 15s 179ms/step - loss: 3.3823e-04 - mae: 0.0120 - acc: 3.7864e-04 Epoch 20/100 83/83 [==============================] - 15s 179ms/step - loss: 2.7471e-04 - mae: 0.0105 - acc: 3.7864e-04 Epoch 21/100 83/83 [==============================] - 15s 184ms/step - loss: 3.1111e-04 - mae: 0.0114 - acc: 3.7864e-04 Epoch 22/100 83/83 [==============================] - 15s 182ms/step - loss: 2.7048e-04 - mae: 0.0108 - acc: 3.7864e-04 Epoch 23/100 83/83 [==============================] - 15s 177ms/step - loss: 2.6737e-04 - mae: 0.0106 - acc: 3.7864e-04 Epoch 24/100 83/83 [==============================] - 12s 143ms/step - loss: 3.0102e-04 - mae: 0.0114 - acc: 3.7864e-04 Epoch 25/100 83/83 [==============================] - 12s 148ms/step - loss: 2.7399e-04 - mae: 0.0109 - acc: 3.7864e-04 Epoch 26/100 83/83 [==============================] - 12s 143ms/step - loss: 2.8185e-04 - mae: 0.0110 - acc: 3.7864e-04 Epoch 27/100 83/83 [==============================] - 12s 143ms/step - loss: 2.7066e-04 - mae: 0.0108 - acc: 3.7864e-04 Epoch 28/100 83/83 [==============================] - 12s 141ms/step - loss: 2.4524e-04 - mae: 0.0104 - acc: 3.7864e-04 Epoch 29/100 83/83 [==============================] - 12s 140ms/step - loss: 2.4877e-04 - mae: 0.0104 - acc: 3.7864e-04 Epoch 30/100 83/83 [==============================] - 12s 140ms/step - loss: 2.5145e-04 - mae: 0.0108 - acc: 3.7864e-04 Epoch 31/100 83/83 [==============================] - 12s 142ms/step - loss: 2.5053e-04 - mae: 0.0104 - acc: 3.7864e-04 Epoch 32/100 83/83 [==============================] - 12s 144ms/step - loss: 2.5941e-04 - mae: 0.0107 - acc: 3.7864e-04 Epoch 33/100 83/83 [==============================] - 12s 145ms/step - loss: 2.1628e-04 - mae: 0.0096 - acc: 3.7864e-04 Epoch 34/100 83/83 [==============================] - 12s 148ms/step - loss: 2.5347e-04 - mae: 0.0104 - acc: 3.7864e-04 Epoch 35/100 83/83 [==============================] - 12s 144ms/step - loss: 2.8450e-04 - mae: 0.0117 - acc: 3.7864e-04 Epoch 36/100 83/83 [==============================] - 12s 147ms/step - loss: 2.1696e-04 - mae: 0.0097 - acc: 3.7864e-04 Epoch 37/100 83/83 [==============================] - 12s 141ms/step - loss: 2.1587e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 38/100 83/83 [==============================] - 12s 144ms/step - loss: 2.1839e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 39/100 83/83 [==============================] - 12s 145ms/step - loss: 2.1115e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 40/100 83/83 [==============================] - 12s 142ms/step - loss: 1.9515e-04 - mae: 0.0094 - acc: 3.7864e-04 Epoch 41/100 83/83 [==============================] - 12s 146ms/step - loss: 2.3016e-04 - mae: 0.0102 - acc: 3.7864e-04 Epoch 42/100 83/83 [==============================] - 12s 142ms/step - loss: 2.1981e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 43/100 83/83 [==============================] - 12s 143ms/step - loss: 2.2498e-04 - mae: 0.0102 - acc: 3.7864e-04 Epoch 44/100 83/83 [==============================] - 12s 143ms/step - loss: 2.5379e-04 - mae: 0.0109 - acc: 3.7864e-04 Epoch 45/100 83/83 [==============================] - 12s 143ms/step - loss: 2.1705e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 46/100 83/83 [==============================] - 12s 140ms/step - loss: 2.2145e-04 - mae: 0.0103 - acc: 3.7864e-04 Epoch 47/100 83/83 [==============================] - 12s 139ms/step - loss: 1.9548e-04 - mae: 0.0092 - acc: 3.7864e-04 Epoch 48/100 83/83 [==============================] - 12s 141ms/step - loss: 2.0833e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 49/100 83/83 [==============================] - 12s 141ms/step - loss: 1.9410e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 50/100 83/83 [==============================] - 12s 142ms/step - loss: 1.9222e-04 - mae: 0.0094 - acc: 3.7864e-04 Epoch 51/100 83/83 [==============================] - 12s 139ms/step - loss: 2.2112e-04 - mae: 0.0103 - acc: 3.7864e-04 Epoch 52/100 83/83 [==============================] - 12s 141ms/step - loss: 1.9559e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 53/100 83/83 [==============================] - 12s 142ms/step - loss: 2.0208e-04 - mae: 0.0096 - acc: 3.7864e-04 Epoch 54/100 83/83 [==============================] - 12s 141ms/step - loss: 1.8062e-04 - mae: 0.0092 - acc: 3.7864e-04 Epoch 55/100 83/83 [==============================] - 12s 142ms/step - loss: 1.9474e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 56/100 83/83 [==============================] - 12s 141ms/step - loss: 2.0758e-04 - mae: 0.0096 - acc: 3.7864e-04 Epoch 57/100 83/83 [==============================] - 12s 142ms/step - loss: 1.9136e-04 - mae: 0.0093 - acc: 3.7864e-04 Epoch 58/100 83/83 [==============================] - 12s 141ms/step - loss: 2.1050e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 59/100 83/83 [==============================] - 12s 141ms/step - loss: 2.2446e-04 - mae: 0.0104 - acc: 3.7864e-04 Epoch 60/100 83/83 [==============================] - 12s 142ms/step - loss: 2.1025e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 61/100 83/83 [==============================] - 12s 142ms/step - loss: 1.9358e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 62/100 83/83 [==============================] - 12s 143ms/step - loss: 1.9089e-04 - mae: 0.0094 - acc: 3.7864e-04 Epoch 63/100 83/83 [==============================] - 12s 143ms/step - loss: 2.0180e-04 - mae: 0.0097 - acc: 3.7864e-04 Epoch 64/100 83/83 [==============================] - 12s 143ms/step - loss: 1.9502e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 65/100 83/83 [==============================] - 12s 142ms/step - loss: 1.9852e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 66/100 83/83 [==============================] - 12s 141ms/step - loss: 2.0266e-04 - mae: 0.0100 - acc: 3.7864e-04 Epoch 67/100 83/83 [==============================] - 12s 143ms/step - loss: 1.8904e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 68/100 83/83 [==============================] - 12s 142ms/step - loss: 2.0111e-04 - mae: 0.0097 - acc: 3.7864e-04 Epoch 69/100 83/83 [==============================] - 12s 144ms/step - loss: 1.9283e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 70/100 83/83 [==============================] - 12s 146ms/step - loss: 1.9512e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 71/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7657e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 72/100 83/83 [==============================] - 12s 145ms/step - loss: 1.7844e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 73/100 83/83 [==============================] - 12s 144ms/step - loss: 2.1166e-04 - mae: 0.0098 - acc: 3.7864e-04 Epoch 74/100 83/83 [==============================] - 12s 145ms/step - loss: 1.6239e-04 - mae: 0.0088 - acc: 3.7864e-04 Epoch 75/100 83/83 [==============================] - 12s 145ms/step - loss: 2.0147e-04 - mae: 0.0097 - acc: 3.7864e-04 Epoch 76/100 83/83 [==============================] - 12s 143ms/step - loss: 1.8721e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 77/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7040e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 78/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7755e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 79/100 83/83 [==============================] - 12s 143ms/step - loss: 1.8537e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 80/100 83/83 [==============================] - 12s 143ms/step - loss: 1.7636e-04 - mae: 0.0089 - acc: 3.7864e-04 Epoch 81/100 83/83 [==============================] - 12s 142ms/step - loss: 1.7798e-04 - mae: 0.0092 - acc: 3.7864e-04 Epoch 82/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7796e-04 - mae: 0.0092 - acc: 3.7864e-04 Epoch 83/100 83/83 [==============================] - 12s 145ms/step - loss: 1.7401e-04 - mae: 0.0090 - acc: 3.7864e-04 Epoch 84/100 83/83 [==============================] - 12s 145ms/step - loss: 2.0091e-04 - mae: 0.0096 - acc: 3.7864e-04 Epoch 85/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7121e-04 - mae: 0.0089 - acc: 3.7864e-04 Epoch 86/100 83/83 [==============================] - 12s 144ms/step - loss: 1.8250e-04 - mae: 0.0092 - acc: 3.7864e-04 Epoch 87/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7934e-04 - mae: 0.0094 - acc: 3.7864e-04 Epoch 88/100 83/83 [==============================] - 12s 145ms/step - loss: 1.6739e-04 - mae: 0.0087 - acc: 3.7864e-04 Epoch 89/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7841e-04 - mae: 0.0090 - acc: 3.7864e-04 Epoch 90/100 83/83 [==============================] - 12s 144ms/step - loss: 1.7798e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 91/100 83/83 [==============================] - 12s 145ms/step - loss: 2.0648e-04 - mae: 0.0099 - acc: 3.7864e-04 Epoch 92/100 83/83 [==============================] - 12s 144ms/step - loss: 1.6237e-04 - mae: 0.0087 - acc: 3.7864e-04 Epoch 93/100 83/83 [==============================] - 12s 144ms/step - loss: 1.5727e-04 - mae: 0.0086 - acc: 3.7864e-04 Epoch 94/100 83/83 [==============================] - 12s 146ms/step - loss: 1.7447e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 95/100 83/83 [==============================] - 12s 144ms/step - loss: 1.9625e-04 - mae: 0.0095 - acc: 3.7864e-04 Epoch 96/100 83/83 [==============================] - 12s 144ms/step - loss: 1.6939e-04 - mae: 0.0089 - acc: 3.7864e-04 Epoch 97/100 83/83 [==============================] - 12s 145ms/step - loss: 1.6023e-04 - mae: 0.0084 - acc: 3.7864e-04 Epoch 98/100 83/83 [==============================] - 12s 143ms/step - loss: 1.7082e-04 - mae: 0.0088 - acc: 3.7864e-04 Epoch 99/100 83/83 [==============================] - 12s 143ms/step - loss: 1.7704e-04 - mae: 0.0091 - acc: 3.7864e-04 Epoch 100/100 83/83 [==============================] - 12s 142ms/step - loss: 1.7186e-04 - mae: 0.0090 - acc: 3.7864e-04
# Getting the predicted stock price
test_data = training_scaled_data[training_data_len - 180: , : ]
#Create the x_test and y_test data sets
X_test = []
y_test = target[training_data_len : , : ]
for i in range(180,len(test_data)):
X_test.append(test_data[i-180:i,0])
# Convert x_test to a numpy array
X_test = array(X_test)
#Reshape the data into the shape accepted by the LSTM
X_test = reshape(X_test, (X_test.shape[0],X_test.shape[1],1))
print('Number of rows and columns: ', X_test.shape)# Making predictions using the test dataset
predicted_stock_price = model.predict(X_test)
predicted_stock_price = sc.inverse_transform(predicted_stock_price)# Visualising the results
train = data_target[:training_data_len]
valid = data_target[training_data_len:]
valid['Predictions'] = predicted_stock_price
plt.figure(figsize=(10,5))
plt.title('Model')
plt.xlabel('Date', fontsize=8)
plt.ylabel('Close Price', fontsize=12)
plt.plot(train['Close'])
plt.plot(valid[['Close', 'Predictions']])
plt.legend(['Train', 'Val', 'Predictions'], loc='lower right')
plt.show()