Pandas 操作CSV

最近项目需求做ML,我对这个又不是很熟只有一点一点学习,其中包括Data的预处理,会用到操作csv,顺便总结下.

Install pandas

pip install pandas

Use

我使用的demo csv如下

读取csv数据并打印

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    print(data)

describe()方法数据统计

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    print(data.describe())

读取文件前几行

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    print(data.head(5))

读取某几行数据

其中loc 与 iloc的区别可以看这篇blog https://blog.csdn.net/u014712482/article/details/85080864
简而言之,loc的参数是具体的列名字行名或者范围,而iloc是行的index与列的index或者范围

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    print(data.loc[2:5])

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    print(data.iloc[2:5,0:3])

选取条件行

df=file.loc[(file['cs-uri'] == url) & (file['cs-method'] == method)]

修改数据并写入

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    data.loc[:,'CO2']=28
    data.to_csv("./targetCsv.csv")

import pandas  as pd

path="./cars.csv"

with open(path)as file:
    data=pd.read_csv(file)
    data.loc[3,'Car']='Test Write'
    data.to_csv("./targetCsv.csv")

创建空csv并写入数据

import pandas  as pd

df=pd.DataFrame({'Yes': [50, 21] , 'No': [131, 2]}, index = ['A','B'])
df.to_csv('./targetCsv.csv')

或者这样

import pandas  as pd

df=pd.DataFrame([[35, 21], [41, 34]],
                         columns=['Apples', 'Bananas'],
                              index=['2017 Sales', '2018 Sales'])
df.to_csv('./targetCsv.csv')

读取excel

import pandas as pd 
data = pd.read_excel(filepath,index=False)

操作方式和csv差不多就不具体演示了