-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathStockPredict.py
170 lines (135 loc) · 5.16 KB
/
StockPredict.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
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
# -*- coding: utf-8 -*-
"""
Created on Mon Jan 11 13:13:14 2021
@author: Augus
"""
# encoding=gbk
import pandas as pd
import matplotlib.pyplot as plt
from sklearn import preprocessing
import numpy as np
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense,LSTM,Dropout
import talib
from jqdatasdk import *
from jqdatasdk import finance#获取外盘数据
import datetime
import dateutil
GLOBALTIME=datetime.datetime.now().strftime('%Y-%m-%d')#计算出当前时间
auth('13633411298', 'Jk12345678')
df = get_price('000016.XSHG', start_date='2020-01-01', end_date=GLOBALTIME,
frequency='daily',
#frequency='minute',
fields=['open', 'close', 'high', 'low', 'volume', 'pre_close'])
df['pct_chg'] = ( df['close'] - df['pre_close'])/df['pre_close']
df = df.rename({'volume':'vol'}, axis='columns')#，这种方法比较灵活。
df=df.drop('pre_close',axis=1)
# 各个指标
data=pd.DataFrame(df[['open', 'high', 'close', 'low', 'vol', 'pct_chg']])
## 名称：随机指标,俗称KD
data['slowk'], data['slowd'] = \
talib.STOCH(df['high'].values, df['low'].values, df['close'].values,
fastk_period=9, slowk_period=3, slowk_matype=0, slowd_period=3, slowd_matype=0)
## 名称： 布林线指标
data["BBANDS_upper"],data["BBANDS_middle"],data["BBANDS_lower"] =talib.BBANDS(df['close'].values, matype=talib.MA_Type.T3)
## 名称：平滑异同移动平均线
data["MACD"],df['MACDsignal'],df['MACDhist'] =talib.MACD(df['close'].values, fastperiod=12, slowperiod=26, signalperiod=9)
## 名称：真实波动幅度均值
data["ATR"] = talib.ATR(df['high'].values,df['low'].values, df['close'].values, timeperiod=14)
## AD - Chaikin A/D Line 量价指标
data["AD"] = talib.AD(df['high'].values,df['low'].values, df['close'].values, df['vol'].values)
data=data[~data["slowk"].isin(['NaN'])]
data=data[~data["slowd"].isin(['NaN'])]
data=data[~data["BBANDS_upper"].isin(['NaN'])]
data=data[~data["MACD"].isin(['NaN'])]
data=data[~data["BBANDS_middle"].isin(['NaN'])]
data=data[~data["BBANDS_lower"].isin(['NaN'])]
data=data[~data["ATR"].isin(['NaN'])]
data=data[~data["AD"].isin(['NaN'])]
data=data[['open', 'high', 'close', 'low', 'vol', 'pct_chg','slowk','BBANDS_upper','BBANDS_lower','MACD','slowd']]
data.head(10)
Alldata=data.shape[0]
#print("数据总量:",Alldata)
first_length=int(Alldata*0.8)
#print("训练数据:",first_length)
second_length=int(Alldata*0.2)
#print("测试数据:",second_length)
# 训练时间长度
time_stamp = 2
train = data[0:first_length + time_stamp]
valid = data[first_length - time_stamp:]
judge_result=valid[time_stamp-1:-1]
judge_result=judge_result['close']
judge_result=np.array(judge_result)
# 归一化
scaler = preprocessing.MinMaxScaler(feature_range=(0, 1))
scaled_data_train = scaler.fit_transform(train)
scaled_data_valid = scaler.fit_transform(valid)
# 训练数据获取
x_train, y_train = [], []
for i in range(time_stamp, len(train)):
x_train.append(scaled_data_train[i - time_stamp:i])
y_train.append(scaled_data_train[i, 2])
x_train, y_train = np.array(x_train), np.array(y_train)
# 验证数据获取
x_valid, y_valid = [], []
for i in range(time_stamp, len(valid)):
x_valid.append(scaled_data_valid[i - time_stamp:i])
y_valid.append(scaled_data_valid[i, 2])
x_valid, y_valid = np.array(x_valid), np.array(y_valid)
# lstm训练
epochs = 10
batch_size = 16
model = Sequential()
#model.add(LSTM(units=80, return_sequences=True, input_dim=x_train.shape[-1], input_length=x_train.shape[1]))
#model.add(Dropout(0.2))
model.add(LSTM(units=80, return_sequences=True, input_shape=(x_train.shape[1], x_train.shape[-1])))
model.add(LSTM(units=20))#20
model.add(Dense(units=1))
model.compile(loss='mean_squared_error', optimizer='adam')
history = model.fit(x_train, y_train, epochs=epochs, batch_size=batch_size, verbose=1)
# 模型预测
closing_price = model.predict(x_valid)
scaler.fit_transform(pd.DataFrame(valid['close'].values))
closing_price = scaler.inverse_transform(closing_price)
y_valid = scaler.inverse_transform([y_valid])
total_count=y_valid.size
predict=closing_price.flatten()#np.resize(predict,(-1,))
real=y_valid.flatten()#np.resize(real,(-1,))
#total_count=646
result_sum=0
for i in range(total_count):
if ( predict[i] - judge_result[i] ) * ( real[i] - judge_result[i] ) > 0:
result_sum = result_sum + 1
#print( result_sum/total_count )
#total_count=646
result=[]
for i in range(total_count):
if ( predict[i]> judge_result[i] ) :
result.append(1)
else:
result.append(0)
print("当前时间：", GLOBALTIME)
print("收盘价预测结果：",predict[-1] )
print("前日收盘价：",judge_result[-1] )
print("二分预测结果：")
if result[-1]:
print("涨")
else:
print("跌")
# 画图
plt.figure(figsize=(16, 8))
predict_data = {
'Predictions': closing_price.reshape(1,-1)[0],
'Close': y_valid[0],
'pre_close' : judge_result
}
predict_data = pd.DataFrame(predict_data)
plt.plot(predict_data[['Predictions']],label='Prediction')
plt.plot(predict_data[['Close']],label='Real')
plt.plot(predict_data[['pre_close']],label='Close')
plt.xlabel('time')
plt.ylabel('price')
plt.legend()
plt.savefig('predict.png')
plt.show()