Python与数据科学实战课程——第三章pandas：Dataframe

import numpy as np
import pandas as pd
from pandas import Series,DataFrame

生成DataFrame

import webbrowser
link ="https://www.tiobe.com/tiobe-index/"
webbrowser.open(link)     #打开一个网站，并从该网站中复制数据

True

df = pd.read_clipboard()  #从粘贴板中获取数据并解析，转换成个dataframe
df

	Aug 2020	Aug 2019	Change	Programming Language	Ratings	Change.1
0	1	2		C	16.98%	+1.83%
1	2	1		Java	14.43%	-1.60%
2	3	3	NaN	Python	9.69%	-0.33%
3	4	4	NaN	C++	6.84%	+0.78%
4	5	5	NaN	C#	4.68%	+0.83%
5	6	6	NaN	Visual Basic	4.66%	+0.97%
6	7	7	NaN	JavaScript	2.87%	+0.62%
7	8	20		R	2.79%	+1.97%
8	9	8		PHP	2.24%	+0.17%

type(df)

pandas.core.frame.DataFrame

dataframe相关属性

df.columns         #获取列名

Index([‘Aug 2020’, ‘Aug 2019’, ‘Change’, ‘Programming Language’, ‘Ratings’,
‘Change.1’],
dtype=‘object’)

df.Ratings            #返回某一列的数值  df.列名

0 16.98%
1 14.43%
2 9.69%
3 6.84%
4 4.68%
5 4.66%
6 2.87%
7 2.79%
8 2.24%
Name: Ratings, dtype: object

df_new = DataFrame(df, columns=["Aug 2020","Programming Language"])  #通过选取需要的列生成新的dataframe
df_new

	Aug2020	ProgrammingLanguage
0	1	C
1	2	Java
2	3	Python
3	4	C++
4	5	C#
5	6	VisualBasic
6	7	JavaScript
7	8	R
8	9	PHP

df["Aug 2020"]

0 1
1 2
2 3
3 4
4 5
5 6
6 7
7 8
8 9
Name: Aug 2020, dtype: int64

type(df["Aug 2020"])        #可以看到dataframe的某一列的数据结构是series

pandas.core.series.Series

df_new = DataFrame(df, columns=["Aug 2020","Programming Language","Sep 2020"])
df_new  #可以看到选取原有dataframe中没有的列并不会报错，而是系统自动生成该列，只是数值都为NaN

	Aug 2020	Programming Language	Sep 2020
0	1	C	NaN
1	2	Java	NaN
2	3	Python	NaN
3	4	C++	NaN
4	5	C#	NaN
5	6	Visual Basic	NaN
6	7	JavaScript	NaN
7	8	R	NaN
8	9	PHP	NaN

给dataframe赋值

方法一:使用列表赋值

df_new["Sep 2020"] = range(0,9)
df_new

	Aug 2020	Programming Language	Sep 2020
0	1	C	0
1	2	Java	1
2	3	Python	2
3	4	C++	3
4	5	C#	4
5	6	Visual Basic	5
6	7	JavaScript	6
7	8	R	7
8	9	PHP	8

方法二：使用array赋值

df_new["Sep 2020"] = np.arange(1,10)
df_new

	Aug 2020	Programming Language	Sep 2020
0	1	C	1
1	2	Java	2
2	3	Python	3
3	4	C++	4
4	5	C#	5
5	6	Visual Basic	6
6	7	JavaScript	7
7	8	R	8
8	9	PHP	9

方法三：使用series进行赋值

df_new["Sep 2020"]=pd.Series(np.arange(2,11))
df_new

	Aug 2020	Programming Language	Sep 2020
0	1	C	2
1	2	Java	3
2	3	Python	4
3	4	C++	5
4	5	C#	6
5	6	Visual Basic	7
6	7	JavaScript	8
7	8	R	9
8	9	PHP	10

对dataframe进行部分赋值

df_new["Sep 2020"] = pd.Series([100,200],index=[1,2])
df_new              #可以看到虽然是部分赋值 但是仍然是一整列赋值，只不过没有赋值的地方自动填充为NaN

	Aug 2020	Programming Language	Sep 2020
0	1	C	NaN
1	2	Java	100.0
2	3	Python	200.0
3	4	C++	NaN
4	5	C#	NaN
5	6	Visual Basic	NaN
6	7	JavaScript	NaN
7	8	R	NaN
8	9	PHP	NaN