Basic

import pandas as pd
print (f" Using {pd.__name__},Version {pd.__version__}")

 Using pandas,Version 0.23.0

创建空Dataframe

df = pd.DataFrame() 
print(df)

Empty DataFrame
Columns: []
Index: []

从Dict创建Dataframe

dict = {'name':["Tom", "Bob", "Mary", "James"], 
        'age': [18, 30, 25, 40], 
        'city':["Beijing", "ShangHai","GuangZhou", "ShenZhen"]} 
  
df = pd.DataFrame(dict) 
df

	name	age	city
0	Tom	18	Beijing
1	Bob	30	ShangHai
2	Mary	25	GuangZhou
3	James	40	ShenZhen

index = pd.Index(["Tom", "Bob", "Mary", "James"],name = 'person')
cols = ['age','city']
data = [[18,'Beijing'],
        [30,'ShangHai'],
        [25,'GuangZhou'],
        [40,'ShenZhen']]

df =pd.DataFrame(index = index,data =data,columns = cols)
df

	age	city
person
Tom	18	Beijing
Bob	30	ShangHai
Mary	25	GuangZhou
James	40	ShenZhen

对columns的基础操作

add column

dict = {'name':["Tom", "Bob", "Mary", "James"], 
        'age': [18, 30, 25, 40], 
        'city':["Beijing", "ShangHai","GuangZhou", "ShenZhen"]} 
  
df = pd.DataFrame(dict) 
df

	name	age	city
0	Tom	18	Beijing
1	Bob	30	ShangHai
2	Mary	25	GuangZhou
3	James	40	ShenZhen

df['country'] = 'USA'
df

	name	age	city	country
0	Tom	18	Beijing	USA
1	Bob	30	ShangHai	USA
2	Mary	25	GuangZhou	USA
3	James	40	ShenZhen	USA

df['adress'] = df['country']
df

	name	age	city	country	adress
0	Tom	18	Beijing	USA	USA
1	Bob	30	ShangHai	USA	USA
2	Mary	25	GuangZhou	USA	USA
3	James	40	ShenZhen	USA	USA

Change column values

df['country'] = 'China'
df

	name	age	city	country	adress
0	Tom	18	Beijing	China	USA
1	Bob	30	ShangHai	China	USA
2	Mary	25	GuangZhou	China	USA
3	James	40	ShenZhen	China	USA

df['adress'] = df['city']+','+ df['country']
df

	name	age	city	country	adress
0	Tom	18	Beijing	China	Beijing,China
1	Bob	30	ShangHai	China	ShangHai,China
2	Mary	25	GuangZhou	China	GuangZhou,China
3	James	40	ShenZhen	China	ShenZhen,China

Delete columns

df.drop('country',axis=1, inplace=True)
del df['city']
df

	name	age	adress
0	Tom	18	Beijing,China
1	Bob	30	ShangHai,China
2	Mary	25	GuangZhou,China
3	James	40	ShenZhen,China

Select columns

df['age']

0    18
1    30
2    25
3    40
Name: age, dtype: int64

df.name

0      Tom
1      Bob
2     Mary
3    James
Name: name, dtype: object

df[['age','name']]  

	age	name
0	18	Tom
1	30	Bob
2	25	Mary
3	40	James

df.columns

Index(['name', 'age', 'adress'], dtype='object')

Rename columns

df.rename(index = str, columns = {'age':'Age','name':'Name','adress':'Adress'},inplace=True)
df

	Name	Age	Adress
0	Tom	18	Beijing,China
1	Bob	30	ShangHai,China
2	Mary	25	GuangZhou,China
3	James	40	ShenZhen,China

df.rename(str.lower, axis='columns',inplace =True)
df

	name	age	adress
0	Tom	18	Beijing,China
1	Bob	30	ShangHai,China
2	Mary	25	GuangZhou,China
3	James	40	ShenZhen,China

df.rename(str.capitalize, axis='columns',inplace =True)
df

	Name	Age	Adress
0	Tom	18	Beijing,China
1	Bob	30	ShangHai,China
2	Mary	25	GuangZhou,China
3	James	40	ShenZhen,China

Set column value with conditions

df['Group'] = 'elderly'
df.loc[df['Age']<=18, 'Group'] = 'young'
df.loc[(df['Age'] >18) & (df['Age'] <= 30), 'Group'] = 'middle_aged'
df

	Name	Age	Adress	Group
0	Tom	18	Beijing,China	young
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly

对rows的基础操作

loc函数查询

df

	Name	Age	Adress	Group
0	Tom	18	Beijing,China	young
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly

df.loc[:]

	Name	Age	Adress	Group
0	Tom	18	Beijing,China	young
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly

loc函数条件查询

df.loc[df['Age']>20]

	Name	Age	Adress	Group
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly

loc函数条件行列查询

df.loc[df['Group']=='middle_aged','Name']

1     Bob
2    Mary
Name: Name, dtype: object

Where 查询

filter_adult = df['Age']>25
result = df.where(filter_adult)
result

	Name	Age	Adress	Group
0	NaN	NaN	NaN	NaN
1	Bob	30.0	ShangHai,China	middle_aged
2	NaN	NaN	NaN	NaN
3	James	40.0	ShenZhen,China	elderly

Query 筛选

df

	Name	Age	Adress	Group
0	Tom	18	Beijing,China	young
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly

df.query('Group=="middle_aged"'and 'Age>30' )

	Name	Age	Adress	Group
3	James	40	ShenZhen,China	elderly

Dataframe其他信息

df.shape

(4, 4)

df.describe()

	Age
count	4.000000
mean	28.250000
std	9.251126
min	18.000000
25%	23.250000
50%	27.500000
75%	32.500000
max	40.000000

df.head(3)
df.tail(3)

	Name	Age	Adress	Group
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly

读写CSV

把df导出为CSV，不要index

df.to_csv('person.csv',index=None,sep=',')

读取CSV为dataframe

person = pd.read_csv('person.csv')
person

	Name	Age	Adress	Group
0	Tom	18	Beijing,China	young
1	Bob	30	ShangHai,China	middle_aged
2	Mary	25	GuangZhou,China	middle_aged
3	James	40	ShenZhen,China	elderly