Python based on Baidu intelligent cloud platform stock information sentiment analysis

Python based on Baidu intelligent cloud platform stock information sentiment analysis

All codes and data addresses are as follows: Python based on Baidu Smart Cloud Platform Stock Information Sentiment Analysis

---

本文章详细内容如下：

---

import the corresponding package

1. Import library

import pandas as pd
import datetime
from aip import AipNlp
import codecs

2. Set account secret key

APP_ID = ''
API_KEY = ''
SECRET_KEY = ''

client = AipNlp(APP_ID, API_KEY, SECRET_KEY)

3. Import data

df = pd.read_csv('600519.csv')
del df["Unnamed: 0"]
df["date"] = pd.to_datetime(df["create time"].apply(lambda x: "2023-"+ x[:5]))

data= df[["date","title"]].sort_values(by="date").reset_index(drop=True)
data

4. Data Merging

newvdata = data.groupby('date').agg(lambda x:list(x)) #按日期分组，把同一天的评论并到一起
newvdata

5. Baidu Natural Language Processing Sentiment Analysis

sitems=client.sentimentClassify(text)['items'][0]#情感分析
   
for i in range(newvdata.shape[0]):
    print('正在处理第{}条,还剩{}条'.format(i,newvdata.shape[0]-1))
    dates=newvdata.index[i]
    for view in newvdata.title[i]:
        print(view)
        get_sentiments(view,dates)

6. Analysis results

sent_df = pd.read_csv("./sentiment.csv",header=None)
sent_df.columns = ["date","positive_prob", "confidence", "sentiment"]
sent_df.date = pd.to_datetime(sent_df.date)
sent_df = sent_df.sort_values(by="date")
group_df = sent_df.groupby('date')["positive_prob"].mean()
group_df = pd.DataFrame(group_df)
group_df

group_df.plot(figsize=(20, 5))

7. Get stock data

import tushare as ts 
# 茅台股票价格数据获取
stock_df = ts.get_k_data('600519',start='2023-01-03',end='2023-07-12')
stock_df["pct_change"] = stock_df["close"].pct_change()*10
stock_df = stock_df.dropna()
stock_df = stock_df.set_index('date')  # 设置日期为索引
stock_df.index = pd.to_datetime(stock_df.index)
stock_df

all_data = stock_df.join(group_df, how="left")
all_data = all_data.fillna(method="ffill")
all_data

all_data[["pct_change", "positive_prob"]].plot(figsize=(20, 5))