import pandas as pd 
df2016=pd.read_csv("https://raw.githubusercontent.com/PacktPublishing/Pandas-Cookbook/master/data/stocks_2016.csv")
df2017=pd.read_csv("https://raw.githubusercontent.com/PacktPublishing/Pandas-Cookbook/master/data/stocks_2017.csv")
df2018=pd.read_csv("https://raw.githubusercontent.com/PacktPublishing/Pandas-Cookbook/master/data/stocks_2018.csv")

pd.concat([df2016,df2017,df2018]) # default는 아래로 쌓음

pd.concat([df2016,df2017,df2018]).reset_index() # 구 인덱스는 아직 존재함

pd.concat([df2016,df2017,df2018]).reset_index(drop=True)

특정 구간만 뽑아서 `concat`하기

pd.concat([df2016,df2017.iloc[:,1:]]) # df2017에서 선택되지 못한 자료인 0열은 자동으로 NaN처리

pd.concat([df2016,df2017,df2018],axis='columns') 
## axis='columns'  axis=1로 가능/ 1대신에 0쓰면 rbind, 또는 그냥 aixs=0 안 써도 자동으로 rbind 처리

df=pd.concat([df2016,df2017,df2018],keys=[2016,2017,2018])
df

df.index

MultiIndex([(2016, 0),
            (2016, 1),
            (2016, 2),
            (2017, 0),
            (2017, 1),
            (2017, 2),
            (2017, 3),
            (2017, 4),
            (2017, 5),
            (2018, 0),
            (2018, 1),
            (2018, 2)],
           )

$MuiliIndex$가 됐음을 알 수 있다.

pd.concat({2016:df2016,2017:df2017,2018:df2018})

pd.concat(dict(zip([2016,2017,2018],[df2016,df2017,df2018])))

pd.concat([df2016,df2017,df2018],keys=[2016,2017])

pd.concat([df2016,df2017,df2018],keys=[2016,2018]) #이렇게 해도 이름만 2018이지 데이터는 df2017로 불러오게 된다

pd.concat([df2016,df2017,df2018],keys=['a','S','C',2019]) #2019는 그냥 아무처리도 못함

`pd.concat([...],keys=[...],axis='columns')`

df=pd.concat([df2016,df2017,df2018],axis=1,keys=[2016,2017,2018])
df

df.columns

MultiIndex([(2016, 'Symbol'),
            (2016, 'Shares'),
            (2016,    'Low'),
            (2016,   'High'),
            (2017, 'Symbol'),
            (2017, 'Shares'),
            (2017,    'Low'),
            (2017,   'High'),
            (2018, 'Symbol'),
            (2018, 'Shares'),
            (2018,    'Low'),
            (2018,   'High')],
           )

column도 MultiIndex로 들어갔음

pd.concat({2016:df2016,2017:df2017,2018:df2018},axis=1)

pd.concat(dict(zip([2016,2017,2018],[df2016,df2017,df2018])),axis=1)

`타이디 데이터`

df=pd.concat(dict(zip([2016,2017,2018],[df2016,df2017,df2018])))
df.stack().reset_index().drop(['level_1'],axis=1).rename(columns={'level_0':'year','level_2':'category',0:'value'})

	Symbol	Shares	Low	High
0	AAPL	80	95	110
1	TSLA	50	80	130
2	WMT	40	55	70
0	AAPL	50	120	140
1	GE	100	30	40
2	IBM	87	75	95
3	SLB	20	55	85
4	TXN	500	15	23
5	TSLA	100	100	300
0	AAPL	40	135	170
1	AMZN	8	900	1125
2	TSLA	50	220	400

	index	Symbol	Shares	Low	High
0	0	AAPL	80	95	110
1	1	TSLA	50	80	130
2	2	WMT	40	55	70
3	0	AAPL	50	120	140
4	1	GE	100	30	40
5	2	IBM	87	75	95
6	3	SLB	20	55	85
7	4	TXN	500	15	23
8	5	TSLA	100	100	300
9	0	AAPL	40	135	170
10	1	AMZN	8	900	1125
11	2	TSLA	50	220	400

	Symbol	Shares	Low	High
0	AAPL	80	95	110
1	TSLA	50	80	130
2	WMT	40	55	70
3	AAPL	50	120	140
4	GE	100	30	40
5	IBM	87	75	95
6	SLB	20	55	85
7	TXN	500	15	23
8	TSLA	100	100	300
9	AAPL	40	135	170
10	AMZN	8	900	1125
11	TSLA	50	220	400

	Symbol	Shares	Low	High
0	AAPL	80	95	110
1	TSLA	50	80	130
2	WMT	40	55	70
0	NaN	50	120	140
1	NaN	100	30	40
2	NaN	87	75	95
3	NaN	20	55	85
4	NaN	500	15	23
5	NaN	100	100	300

	Symbol	Shares	Low	High	Symbol	Shares	Low	High	Symbol	Shares	Low	High
0	AAPL	80.0	95.0	110.0	AAPL	50	120	140	AAPL	40.0	135.0	170.0
1	TSLA	50.0	80.0	130.0	GE	100	30	40	AMZN	8.0	900.0	1125.0
2	WMT	40.0	55.0	70.0	IBM	87	75	95	TSLA	50.0	220.0	400.0
3	NaN	NaN	NaN	NaN	SLB	20	55	85	NaN	NaN	NaN	NaN
4	NaN	NaN	NaN	NaN	TXN	500	15	23	NaN	NaN	NaN	NaN
5	NaN	NaN	NaN	NaN	TSLA	100	100	300	NaN	NaN	NaN	NaN

		Symbol	Shares	Low	High
2016	0	AAPL	80	95	110
	1	TSLA	50	80	130
	2	WMT	40	55	70
2017	0	AAPL	50	120	140
	1	GE	100	30	40
	2	IBM	87	75	95
	3	SLB	20	55	85
	4	TXN	500	15	23
	5	TSLA	100	100	300
2018	0	AAPL	40	135	170
	1	AMZN	8	900	1125
	2	TSLA	50	220	400

특정 구간만 뽑아서 concat하기

pd.concat([...],keys=[...],axis='columns')

타이디 데이터

특정 구간만 뽑아서 `concat`하기

`pd.concat([...],keys=[...],axis='columns')`

`타이디 데이터`