predict_stock_market/analysis.py at master · byander/predict_stock_market · GitHub

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
import pandas as pd
import numpy as np
import plotly.graph_objs as go
from plotly.offline import plot
from sklearn.svm import SVR
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error
from data import get_dataset


def set_training():
    ds = get_dataset()
    features = ['open', 'high', 'low', 'volume']
    training = ds[features]
    y = ds.close

    x_training, x_test, y_training, y_test = train_test_split(training, y, random_state=42)

    # Create model
    lr_model = LinearRegression()
    lr_model.fit(x_training, y_training)

    predict = lr_model.predict(x_test)
    RMSE = mean_squared_error(y_test, predict) ** 0.5

    # Compare real data with predict data
    # print(y_test[:5])
    # print(predict[:5])

    x = ds.date
    real_data = go.Scatter(
        x=x,
        y=y_test,
        # mode='markers',
        name='Preço real'
    )

    predict_data = go.Scatter(
        x=x,
        y=predict,
        name='Preço Estimado'
    )

    layout = go.Layout(
        xaxis=dict(
            title='Data'
        ),
        yaxis=dict(
            range=[x, y],
            title='Valor da ação'
        ),
        showlegend=True
    )
    data = [real_data, predict_data]
    fig = go.Figure(data=data, layout=layout)
    plot(fig, filename='Charts_tmp/training')

    # print(y_test)
    # print(predict)
    # print(RMSE)

def predict_prices():
    ds = get_dataset()
    ds['day'] = pd.to_datetime(ds['date']).dt.day
    days = ['day']
    training = ds[days]

    dates = ds.day
    dates_index = training.set_index(ds.day)
    prices = ds.close

    dates2 = np.reshape(dates_index, (len(dates_index), 1))

    # svr_lin = SVR(kernel='linear', C=1000)
    # svr_poly = SVR(kernel='poly', C=1000, degree=2)
    svr_rbf = SVR(kernel='rbf', C=1000, gamma=0.1)

    # svr_lin.fit(dates2, prices)
    # svr_poly.fit(dates2, prices)
    svr_rbf.fit(dates2, prices)

    real_data = go.Scatter(
        x=ds['day'],
        y=prices,
        mode='markers',
        name='Preço real'
    )
    # lin = go.Scatter(
    #     x=ds['day'],
    #     y=svr_lin.predict(dates2),
    #     name='Linear'
    # )
    # poly = go.Scatter(
    #     x=ds['day'],
    #     y=svr_poly.predict(dates2),
    #     name='Poly'
    # )
    rbf = go.Scatter(
        x=ds['day'],
        y=svr_rbf.predict(dates2),
        name='RBF'
    )

    layout = go.Layout(
        xaxis=dict(
            title='Data'
        ),
        yaxis=dict(
            title='Valor da ação'
        ),
        showlegend=True
    )
    data = [real_data, rbf]
    fig = go.Figure(data=data, layout=layout)
    plot(fig, filename='Charts_tmp/training')


# set_training()
predict_prices()