In [ ]:

import pandas as pd

In [57]:

# Multi-index DataFrame
sales_df = pd.read_csv('Online Retail.csv', encoding = 'unicode_escape')
sales_df.set_index(keys=['Country','InvoiceDate'], inplace = True)
sales_df.sort_index(inplace=True)
sales_df

Out[57]:

		InvoiceNo	StockCode	Description	Quantity	UnitPrice	CustomerID
Country	InvoiceDate
Australia	1/10/2011 9:58	540557	22523	CHILDS GARDEN FORK PINK	96	0.85	12415.0
	1/11/2011 9:47	540700	21581	SKULLS DESIGN COTTON TOTE BAG	6	2.25	12393.0
	1/11/2011 9:47	540700	22619	SET OF 6 SOLDIER SKITTLES	8	3.75	12393.0
	1/11/2011 9:47	540700	84997B	RED 3 PIECE RETROSPOT CUTLERY SET	6	3.75	12393.0
	1/11/2011 9:47	540700	20727	LUNCH BAG BLACK SKULL.	20	1.65	12393.0
...	...	...	...	...	...	...	...
Unspecified	9/2/2011 12:17	565303	85227	SET OF 6 3D KIT CARDS FOR KIDS	4	0.85	NaN
	9/2/2011 12:17	565303	22138	BAKING SET 9 PIECE RETROSPOT	2	4.95	NaN
	9/2/2011 12:17	565303	21889	WOODEN BOX OF DOMINOES	5	1.25	NaN
	9/2/2011 12:17	565303	22550	HOLIDAY FUN LUDO	2	3.75	NaN
	9/2/2011 12:17	565303	23127	FELTCRAFT GIRL NICOLE KIT	1	4.95	NaN

541909 rows × 6 columns

In [58]:

# access specific elements
sales_df.loc['Australia','1/11/2011 9:47']

Out[58]:

		InvoiceNo	StockCode	Description	Quantity	UnitPrice	CustomerID
Country	InvoiceDate
Australia	1/11/2011 9:47	540700	21581	SKULLS DESIGN COTTON TOTE BAG	6	2.25	12393.0
	1/11/2011 9:47	540700	22619	SET OF 6 SOLDIER SKITTLES	8	3.75	12393.0
	1/11/2011 9:47	540700	84997B	RED 3 PIECE RETROSPOT CUTLERY SET	6	3.75	12393.0
	1/11/2011 9:47	540700	20727	LUNCH BAG BLACK SKULL.	20	1.65	12393.0
	1/11/2011 9:47	540700	20726	LUNCH BAG WOODLAND	20	1.65	12393.0
	1/11/2011 9:47	540700	22383	LUNCH BAG SUKI DESIGN	10	1.65	12393.0
	1/11/2011 9:47	540700	21249	WOODLAND HEIGHT CHART STICKERS	6	2.95	12393.0
	1/11/2011 9:47	540700	22378	WALL TIDY RETROSPOT	20	0.85	12393.0
	1/11/2011 9:47	540700	22175	PINK OWL SOFT TOY	12	2.95	12393.0
	1/11/2011 9:47	540700	22176	BLUE OWL SOFT TOY	12	2.95	12393.0
	1/11/2011 9:47	540700	84997C	BLUE 3 PIECE POLKADOT CUTLERY SET	6	3.75	12393.0
	1/11/2011 9:47	540700	20728	LUNCH BAG CARS BLUE	20	1.65	12393.0
	1/11/2011 9:47	540700	22382	LUNCH BAG SPACEBOY DESIGN	20	1.65	12393.0
	1/11/2011 9:47	540700	21915	RED HARMONICA IN BOX	12	1.25	12393.0
	1/11/2011 9:47	540700	22549	PICTURE DOMINOES	12	1.45	12393.0
	1/11/2011 9:47	540700	21578	WOODLAND DESIGN COTTON TOTE BAG	12	2.25	12393.0
	1/11/2011 9:47	540700	21577	SAVE THE PLANET COTTON TOTE BAG	12	2.25	12393.0
	1/11/2011 9:47	540700	22245	HOOK, 1 HANGER ,MAGIC GARDEN	12	0.85	12393.0
	1/11/2011 9:47	540700	22244	3 HOOK HANGER MAGIC GARDEN	12	1.95	12393.0

In [61]:

# first argument references rows and the second argument references a column
# feed index as a tuple (to avoid confusion about arguments)
sales_df.loc[('Australia','1/11/2011 9:47'),'UnitPrice']

Out[61]:

Country    InvoiceDate   
Australia  1/11/2011 9:47    2.25
           1/11/2011 9:47    3.75
           1/11/2011 9:47    3.75
           1/11/2011 9:47    1.65
           1/11/2011 9:47    1.65
           1/11/2011 9:47    1.65
           1/11/2011 9:47    2.95
           1/11/2011 9:47    0.85
           1/11/2011 9:47    2.95
           1/11/2011 9:47    2.95
           1/11/2011 9:47    3.75
           1/11/2011 9:47    1.65
           1/11/2011 9:47    1.65
           1/11/2011 9:47    1.25
           1/11/2011 9:47    1.45
           1/11/2011 9:47    2.25
           1/11/2011 9:47    2.25
           1/11/2011 9:47    0.85
           1/11/2011 9:47    1.95
Name: UnitPrice, dtype: float64

In [62]:

# the numeric indexing
# use one index so no tuples are required
# 첫 행
sales_df.iloc[0]

Out[62]:

InvoiceNo                       540557
StockCode                        22523
Description    CHILDS GARDEN FORK PINK
Quantity                            96
UnitPrice                         0.85
CustomerID                     12415.0
Name: (Australia, 1/10/2011 9:58), dtype: object

In [63]:

# rows, columns with numeric indexing
sales_df.iloc[0,0]

Out[63]:

'540557'

In [64]:

# transpose indexes and columns
sales_df = sales_df.transpose()
sales_df.head(10)

Out[64]:

Country	Australia										...	Unspecified
InvoiceDate	1/10/2011 9:58	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	1/11/2011 9:47	...	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17	9/2/2011 12:17
InvoiceNo	540557	540700	540700	540700	540700	540700	540700	540700	540700	540700	...	565303	565303	565303	565303	565303	565303	565303	565303	565303	565303
StockCode	22523	21581	22619	84997B	20727	20726	22383	21249	22378	22175	...	20983	20977	20982	23084	22549	85227	22138	21889	22550	23127
Description	CHILDS GARDEN FORK PINK	SKULLS DESIGN COTTON TOTE BAG	SET OF 6 SOLDIER SKITTLES	RED 3 PIECE RETROSPOT CUTLERY SET	LUNCH BAG BLACK SKULL.	LUNCH BAG WOODLAND	LUNCH BAG SUKI DESIGN	WOODLAND HEIGHT CHART STICKERS	WALL TIDY RETROSPOT	PINK OWL SOFT TOY	...	12 PENCILS TALL TUBE RED RETROSPOT	36 PENCILS TUBE WOODLAND	12 PENCILS TALL TUBE SKULLS	RABBIT NIGHT LIGHT	PICTURE DOMINOES	SET OF 6 3D KIT CARDS FOR KIDS	BAKING SET 9 PIECE RETROSPOT	WOODEN BOX OF DOMINOES	HOLIDAY FUN LUDO	FELTCRAFT GIRL NICOLE KIT
Quantity	96	6	8	6	20	20	10	6	20	12	...	4	4	5	1	1	4	2	5	2	1
UnitPrice	0.85	2.25	3.75	3.75	1.65	1.65	1.65	2.95	0.85	2.95	...	0.85	1.25	0.85	2.08	1.45	0.85	4.95	1.25	3.75	4.95
CustomerID	12415.0	12393.0	12393.0	12393.0	12393.0	12393.0	12393.0	12393.0	12393.0	12393.0	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

6 rows × 541909 columns

In [67]:

sales_df.loc['UnitPrice',('Australia','1/10/2011 9:58')]

Out[67]:

Country    InvoiceDate   
Australia  1/10/2011 9:58    0.85
Name: UnitPrice, dtype: object

In [68]:

sales_df.loc['UnitPrice',('Australia',"1/10/2011 9:58"):('Belgium',"1/10/2011 9:58")]

Out[68]:

Country    InvoiceDate   
Australia  1/10/2011 9:58    0.85
           1/11/2011 9:47    2.25
           1/11/2011 9:47    3.75
           1/11/2011 9:47    3.75
           1/11/2011 9:47    1.65
                             ... 
Bahrain    5/9/2011 13:49    2.95
           5/9/2011 13:49    2.95
           5/9/2011 13:49    4.25
           5/9/2011 13:49    4.25
           5/9/2011 13:49    4.25
Name: UnitPrice, Length: 1679, dtype: object

In [69]:

sales_df = pd.read_csv('Online Retail.csv', encoding = 'unicode_escape')
sales_df.set_index(keys=['Country','InvoiceDate'], inplace = True)

In [71]:

# indexes are swopped
sales_df = sales_df.swaplevel()
sales_df

Out[71]:

		InvoiceNo	StockCode	Description	Quantity	UnitPrice	CustomerID
Country	InvoiceDate
United Kingdom	12/1/2010 8:26	536365	85123A	WHITE HANGING HEART T-LIGHT HOLDER	6	2.55	17850.0
	12/1/2010 8:26	536365	71053	WHITE METAL LANTERN	6	3.39	17850.0
	12/1/2010 8:26	536365	84406B	CREAM CUPID HEARTS COAT HANGER	8	2.75	17850.0
	12/1/2010 8:26	536365	84029G	KNITTED UNION FLAG HOT WATER BOTTLE	6	3.39	17850.0
	12/1/2010 8:26	536365	84029E	RED WOOLLY HOTTIE WHITE HEART.	6	3.39	17850.0
...	...	...	...	...	...	...	...
France	12/9/2011 12:50	581587	22613	PACK OF 20 SPACEBOY NAPKINS	12	0.85	12680.0
	12/9/2011 12:50	581587	22899	CHILDREN'S APRON DOLLY GIRL	6	2.10	12680.0
	12/9/2011 12:50	581587	23254	CHILDRENS CUTLERY DOLLY GIRL	4	4.15	12680.0
	12/9/2011 12:50	581587	23255	CHILDRENS CUTLERY CIRCUS PARADE	4	4.15	12680.0
	12/9/2011 12:50	581587	22138	BAKING SET 9 PIECE RETROSPOT	3	4.95	12680.0

541909 rows × 6 columns

[Pandas][DataFrame][MultiIndex]S4_05_Multi_indexing_operations2