-
-
Notifications
You must be signed in to change notification settings - Fork 102
/
stock_prediction_numpy.py
139 lines (111 loc) · 5.23 KB
/
stock_prediction_numpy.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
# Copyright 2020-2024 Jordi Corbilla. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
import os
import numpy as np
from datetime import timedelta
import random
import pandas as pd
from sklearn.preprocessing import MinMaxScaler
from datetime import datetime
import yfinance as yf
class StockData:
def __init__(self, stock):
self._stock = stock
self._sec = yf.Ticker(self._stock.get_ticker())
self._min_max = MinMaxScaler(feature_range=(0, 1))
def __data_verification(self, train):
print('mean:', train.mean(axis=0))
print('max', train.max())
print('min', train.min())
print('Std dev:', train.std(axis=0))
def get_stock_short_name(self):
return self._sec.info['shortName']
def get_min_max(self):
return self._min_max
def get_stock_currency(self):
return self._sec.info['currency']
def download_transform_to_numpy(self, time_steps, project_folder):
end_date = datetime.today()
print('End Date: ' + end_date.strftime("%Y-%m-%d"))
data = yf.download([self._stock.get_ticker()], start=self._stock.get_start_date(), end=end_date)[['Close']]
data = data.reset_index()
data.to_csv(os.path.join(project_folder, 'downloaded_data_'+self._stock.get_ticker()+'.csv'))
#print(data)
training_data = data[data['Date'] < self._stock.get_validation_date()].copy()
test_data = data[data['Date'] >= self._stock.get_validation_date()].copy()
training_data = training_data.set_index('Date')
# Set the data frame index using column Date
test_data = test_data.set_index('Date')
#print(test_data)
train_scaled = self._min_max.fit_transform(training_data)
self.__data_verification(train_scaled)
# Training Data Transformation
x_train = []
y_train = []
for i in range(time_steps, train_scaled.shape[0]):
x_train.append(train_scaled[i - time_steps:i])
y_train.append(train_scaled[i, 0])
x_train, y_train = np.array(x_train), np.array(y_train)
x_train = np.reshape(x_train, (x_train.shape[0], x_train.shape[1], 1))
total_data = pd.concat((training_data, test_data), axis=0)
inputs = total_data[len(total_data) - len(test_data) - time_steps:]
test_scaled = self._min_max.fit_transform(inputs)
# Testing Data Transformation
x_test = []
y_test = []
for i in range(time_steps, test_scaled.shape[0]):
x_test.append(test_scaled[i - time_steps:i])
y_test.append(test_scaled[i, 0])
x_test, y_test = np.array(x_test), np.array(y_test)
x_test = np.reshape(x_test, (x_test.shape[0], x_test.shape[1], 1))
return (x_train, y_train), (x_test, y_test), (training_data, test_data)
def __date_range(self, start_date, end_date):
for n in range(int((end_date - start_date).days)):
yield start_date + timedelta(n)
def negative_positive_random(self):
return 1 if random.random() < 0.5 else -1
def pseudo_random(self):
return random.uniform(0.01, 0.03)
def generate_future_data(self, time_steps, min_max, start_date, end_date, latest_close_price):
x_future = []
y_future = []
# We need to provide a randomisation algorithm for the close price
# This is my own implementation and it will provide a variation of the
# close price for a +-1-3% of the original value, when the value wants to go below
# zero, it will be forced to go up.
original_price = latest_close_price
for single_date in self.__date_range(start_date, end_date):
x_future.append(single_date)
direction = self.negative_positive_random()
random_slope = direction * (self.pseudo_random())
#print(random_slope)
original_price = original_price + (original_price * random_slope)
#print(original_price)
if original_price < 0:
original_price = 0
y_future.append(original_price)
test_data = pd.DataFrame({'Date': x_future, 'Close': y_future})
test_data = test_data.set_index('Date')
test_scaled = min_max.fit_transform(test_data)
x_test = []
y_test = []
#print(test_scaled.shape[0])
for i in range(time_steps, test_scaled.shape[0]):
x_test.append(test_scaled[i - time_steps:i])
y_test.append(test_scaled[i, 0])
#print(i - time_steps)
x_test, y_test = np.array(x_test), np.array(y_test)
x_test = np.reshape(x_test, (x_test.shape[0], x_test.shape[1], 1))
return x_test, y_test, test_data