Python学习10

大家好，这里是七七，今天是Python学习专题的最后一期，要介绍两个代码，一个是名称字符串匹配代码，一个是时间序列分解实现代码。

首先介绍名称字符串匹配代码。

一、名称字符串匹配代码

import pandas as pd
from fuzzywuzzy import fuzz
from fuzzywuzzy import process
import re
info_data=pd.read_excel("./data/附件1.xlsx")
sale_data=pd.read_excel("./data/附件2.xlsx")
buy_data=pd.read_excel("./data/附件3.xlsx")
loss_data=pd.read_excel("./data/附件4.xlsx")

data=pd.merge(buy_data,info_data,on="单品编码",how="left")
data=data[["日期","单品名称"]]
data["日期"]=pd.to_datetime(data["日期"])
data=data.set_index("日期")

grouped=data.groupby("日期")
for group_name,group_data in grouped:
    strings=group_data["单品名称"].tolist()
    threshold=80
    similar_strings={}
    for string in strings:
        best_match=process.extractOne(
            string,
            [s for s in strings if s not in [string]],
            scorer=fuzz.ratio)
        if best_match[1]>=threshold and best_match[0]!=string and best_match[0][:2]==string[:2]:
            if re.search(r'\(\d+\)',best_match[0])and re.search(r'\(\d+\)',string):
                similar_strings[string]=best_match[0]
                strings=[s for s in strings if s not in [string]]
    if bool(similar_strings):
        print(group_name)
    for original,similar in similar_strings.items():
        print(f"主要相同的字符串:'{original}‘和'{similar}'")
print(data.info)

二、时间序列分解实现代码

import matplotlib.pyplot as plt
import pandas as pd
from statsmodels.tsa.seasonal import seasonal_decompose

plt.rcParams['font.sans-serif'] = [u'simHei']
plt.rcParams['axes.unicode_minus'] = False

def time_series_3d(pd_list:list,name):
    num_plots=4
    
    plt.figure(figsize=(8,6))
    
    trend_df=pd.DataFrame()
    
    for df in pd_list:
        
        
        result=seasonal_decompose(df,model='additive',period=365)
        
        for i in range(num_plots):
            if i==1:
                plt.plot(result.trend,label='Trend')
                plt.legend(loc='upper left')
                trend_df[df.name]=result.trend
                
    trend_df.dropna(inplace=True)
    trend_df.to_csv(f"/trend/{name}.csv",encoding="GBK")
    print(trend_df)
    plt.title(name,fontsize=16)
    plt.tight_layout()
    plt.show()
    
######################
#读取数据

info_data=pd.read_excel("./data/附件1.xlsx")
sale_data=pd.read_excel("./data/附件2.xlsx")
buy_data=pd.read_excel("./data/附件3.xlsx")
loss_data=pd.read_excel("./data/附件4.xlsx")

print(sale_data)
sale_data["销售日期"]=pd.to_datetime(sale_data["销售日期"])

rst_data=pd.read_excel("")

####################
#处理

grouped=rst_data.groupby("品类")
for groupe_name,group_data in grouped:
    group_data["销售日期"]=pd.to_datetime(group_data["销售日期"])
    group_data=group_data.set_index("销售日期")
    
    time_series_3d(group_data["销量（千克）"],group_data["利润率"],group_data["批发价格(元/千克)"],group_data["销售单价(元/千克)"],name=groupe_name)