LSTM、GRU、BiLSTM股价预测实例

最新推荐文章于 2023-11-05 01:38:33 发布

经济工科生

最新推荐文章于 2023-11-05 01:38:33 发布

阅读量522

点赞数 1

文章标签： lstm gru python 深度学习 tensorflow

本文链接：https://blog.csdn.net/weixin_49330017/article/details/128928087

版权

【代码】LSTM、GRU、BiLSTM股价预测实例。

摘要由CSDN通过智能技术生成

1、获取股价数据

#安装BaoStock，BaoStock是一个开源的python量化数据接口
pip install baostock
#安装好后，导入baostock包和pandas包
import baostock  as bk
import pandas as pd
#通过以下代码获取股票

lg = bk.login()
# 显示登陆返回信息
print('login respond error_code:'+lg.error_code)
print('login respond  error_msg:'+lg.error_msg)
#### 获取沪深A股历史K线数据 ####
# 详细指标参数和操作，参见Baostock官网http://baostock.com/baostock/
rs = bk.query_history_k_data("sh.601668",
    "date,code,open,high,low,close,volume,turn,pctChg,peTTM,pbMRQ,psTTM,pcfNcfTTM",
    start_date='2010-01-01', end_date='2021-12-31',
    frequency="d", adjustflag="3")
print('query_history_k_data respond error_code:'+rs.error_code)
print('query_history_k_data respond  error_msg:'+rs.error_msg)

#### 打印结果集 ####
data_list = []
while (rs.error_code == '0') & rs.next():
    # 获取一条记录，将记录合并在一起
    data_list.append(rs.get_row_data())
result = pd.DataFrame(data_list, columns=rs.fields)
result. Head()
#### 结果集输出到csv文件 ####   
result.to_csv(r"********\xx.csv", index=False)
#### 登出系统 ####
bk.logout()

2、导入相关库

import os
os.environ["CUDA_VISIBLE_DEVICES"]="-1" #使用CPU来运行
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import keras
from keras.layers import Dense, LSTM, Dropout, GRU, Bidirectional
from keras.callbacks import EarlyStopping, ModelCheckpoint
from keras.models import load_model 
import json
import os
import random
import tensorflow as tf

3、设置随机数种子保证运行结果不变

seed_num = 42 #数字随意
os.environ['PYTHONHASHSEED']=str(seed_num)
random.seed(seed_num)
np.random.seed(seed_num)
tf.random.set_seed(seed_num)
from tensorflow.compat.v1.keras import backend as K
session_conf = tf.compat.v1.ConfigProto(intra_op_parallelism_threads=1,
                             inter_op_parallelism_threads=1)
sess=tf.compat.v1.Session(graph=tf.compat.v1.get_default_graph(),config=session_conf)
K.set_session(sess)

4、设置早停法参数

earlystop = EarlyStopping(monitor='val_loss',
                          min_delta=0,
                          patience=30,
                          verbose=2)

mc = ModelCheckpoint(filepath='best_model.h5',
                     monitor='val_acc',
                     mode='max',
                     verbose=1,
                     save_best_only=True)

5、导入数据和处理数据

data=pd.read_csv(r"*********\xxx.csv")
all_data = data.iloc[:, 2:]
# 特征归一化
from sklearn.preprocessing import MinMaxScaler
sc = MinMaxScaler(feature_range = (0, 1),)
all_data_scaled = sc.fit_transform(all_data)
#划分训练集、验证集、测试集 分别按照80%、10%、10%来划分
features = []
labels = []
a=int(len(all_data_scaled)*0.8)
b=int(len(all_data_scaled)*0.9)
#设置窗口为60天
for i in range(60, len(all_data_scaled)):
    features. Append(all_data_scaled[i-60:i,])
    labels.append(all_data_scaled[i, 1])
features, labels = np.arr