2.Indexing and Slicing In Pandas

Indexing And Slicing In Pandas

Loading Libraries :

In [1]:

import pandas as pd

Loading Dataset :

In [8]:

heroes=pd.read_csv('heroes.csv')
heroes.head()

	sr	name	Gender	Eye color	Race	Hair color	Height	Publisher	Skin color	Alignment	Weight
0	0	A-Bomb	Male	yellow	Human	No Hair	203.0	Marvel Comics	-	good	441.0
1	1	Abe Sapien	Male	blue	Icthyo Sapien	No Hair	191.0	Dark Horse Comics	blue	good	65.0
2	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	DC Comics	red	good	90.0
3	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	Marvel Comics	-	bad	441.0
4	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	Marvel Comics	-	bad	-99.0

We will perform indexing and slicing operation on this Dataset.

First we will have to learn how to set and remove indexes.

Setting and Remove Indexes :

Setting an index :

pandas allows us to set column as an index. Say , if we want to set 'Publisher' as an index.

In [18]:

heroes_publisher = heroes.set_index('Publisher')
heroes_publisher.head()

	sr	name	Gender	Eye color	Race	Hair color	Height	Skin color	Alignment	Weight
Publisher
Marvel Comics	0	A-Bomb	Male	yellow	Human	No Hair	203.0	-	good	441.0
Dark Horse Comics	1	Abe Sapien	Male	blue	Icthyo Sapien	No Hair	191.0	blue	good	65.0
DC Comics	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	red	good	90.0
Marvel Comics	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	-	bad	441.0
Marvel Comics	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	-	bad	-99.0

Resetting index to default :

In [19]:

heroes_default = heroes.reset_index()
heroes_default

	index	sr	name	Gender	Eye color	Race	Hair color	Height	Publisher	Skin color	Alignment	Weight
0	0	0	A-Bomb	Male	yellow	Human	No Hair	203.0	Marvel Comics	-	good	441.0
1	1	1	Abe Sapien	Male	blue	Icthyo Sapien	No Hair	191.0	Dark Horse Comics	blue	good	65.0
2	2	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	DC Comics	red	good	90.0
3	3	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	Marvel Comics	-	bad	441.0
4	4	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	Marvel Comics	-	bad	-99.0
...	...	...	...	...	...	...	...	...	...	...	...	...
729	729	729	Yellowjacket II	Female	blue	Human	Strawberry Blond	165.0	Marvel Comics	-	good	52.0
730	730	730	Ymir	Male	white	Frost Giant	No Hair	304.8	Marvel Comics	white	good	-99.0
731	731	731	Yoda	Male	brown	Yoda's species	White	66.0	George Lucas	green	good	17.0
732	732	732	Zatanna	Female	blue	Human	Black	170.0	DC Comics	-	good	57.0
733	733	733	Zoom	Male	red	-	Brown	185.0	DC Comics	-	bad	81.0

Reset Index and drop previous index:

In [22]:

heroes_publisher.reset_index(drop=True)

	sr	name	Gender	Eye color	Race	Hair color	Height	Skin color	Alignment	Weight
0	0	A-Bomb	Male	yellow	Human	No Hair	203.0	-	good	441.0
1	1	Abe Sapien	Male	blue	Icthyo Sapien	No Hair	191.0	blue	good	65.0
2	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	red	good	90.0
3	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	-	bad	441.0
4	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	-	bad	-99.0
...	...	...	...	...	...	...	...	...	...	...
729	729	Yellowjacket II	Female	blue	Human	Strawberry Blond	165.0	-	good	52.0
730	730	Ymir	Male	white	Frost Giant	No Hair	304.8	white	good	-99.0
731	731	Yoda	Male	brown	Yoda's species	White	66.0	green	good	17.0
732	732	Zatanna	Female	blue	Human	Black	170.0	-	good	57.0
733	733	Zoom	Male	red	-	Brown	185.0	-	bad	81.0

Subsetting with .loc[]

We use .loc[] over the standard square bracket subsetting because .loc[] is convenient, easy to read & understand by other developer. Now, we will see the difference between both the subsetting then you will be able to understand how .loc[] works.

In [24]:

# Making a list of publishers to subset on
publishers=['Marvel Comics','DC Comics']
# Subset heroes using standard square brackets
heroes[heroes['Publisher'].isin(publishers)]

	sr	name	Gender	Eye color	Race	Hair color	Height	Publisher	Skin color	Alignment	Weight
0	0	A-Bomb	Male	yellow	Human	No Hair	203.0	Marvel Comics	-	good	441.0
2	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	DC Comics	red	good	90.0
3	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	Marvel Comics	-	bad	441.0
4	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	Marvel Comics	-	bad	-99.0
5	5	Absorbing Man	Male	blue	Human	No Hair	193.0	Marvel Comics	-	bad	122.0
...	...	...	...	...	...	...	...	...	...	...	...
728	728	Yellowjacket	Male	blue	Human	Blond	183.0	Marvel Comics	-	good	83.0
729	729	Yellowjacket II	Female	blue	Human	Strawberry Blond	165.0	Marvel Comics	-	good	52.0
730	730	Ymir	Male	white	Frost Giant	No Hair	304.8	Marvel Comics	white	good	-99.0
732	732	Zatanna	Female	blue	Human	Black	170.0	DC Comics	-	good	57.0
733	733	Zoom	Male	red	-	Brown	185.0	DC Comics	-	bad	81.0

Using .loc[]

In [28]:

heroes_publisher.loc[publishers]

	sr	name	Gender	Eye color	Race	Hair color	Height	Skin color	Alignment	Weight
Publisher
Marvel Comics	0	A-Bomb	Male	yellow	Human	No Hair	203.0	-	good	441.0
Marvel Comics	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	-	bad	441.0
Marvel Comics	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	-	bad	-99.0
Marvel Comics	5	Absorbing Man	Male	blue	Human	No Hair	193.0	-	bad	122.0
Marvel Comics	8	Agent 13	Female	blue	-	Blond	173.0	-	good	61.0
...	...	...	...	...	...	...	...	...	...	...
DC Comics	715	Wildfire	Male	-	-	-	-99.0	-	good	-99.0
DC Comics	720	Wonder Girl	Female	blue	Demi-God	Blond	165.0	-	good	51.0
DC Comics	722	Wonder Woman	Female	blue	Amazon	Black	183.0	-	good	74.0
DC Comics	732	Zatanna	Female	blue	Human	Black	170.0	-	good	57.0
DC Comics	733	Zoom	Male	red	-	Brown	185.0	-	bad	81.0

Setting multi-level indexes

Indexes can also be made out of multiple columns, forming a multi-level index (sometimes called a hierarchical index). The benefit is that multi-level indexes make it more natural to reason about nested categorical variables. Let's make multi-index in our data and subset it.

Index heroes by Alignment & Publisher

In [29]:

heroes_multi = heroes.set_index(['Alignment','Publisher'])
heroes_multi.head()

		sr	name	Gender	Eye color	Race	Hair color	Height	Skin color	Weight
Alignment	Publisher
good	Marvel Comics	0	A-Bomb	Male	yellow	Human	No Hair	203.0	-	441.0
Dark Horse Comics	1	Abe Sapien	Male	blue	Icthyo Sapien	No Hair	191.0	blue	65.0
DC Comics	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	red	90.0
bad	Marvel Comics	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	-	441.0
Marvel Comics	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	-	-99.0

# List of tuples: neutral, Marvel Comics & neutral, DC Comics
rows_to_keep = [('neutral','Marvel Comics'),('neutral','DC Comics')]

# Subset for rows to keep
heroes_multi.loc[rows_to_keep]

Slicing index values

Slicing lets you select consecutive elements of an object using first:last syntax. DataFrames can be sliced by index values or by row/column number, we will start with the first case. This involves slicing inside the .loc[] method.

In [33]:

heroes_srt=heroes_multi.sort_index()
heroes_srt.head()

		sr	name	Gender	Eye color	Race	Hair color	Height	Skin color	Weight
Alignment	Publisher
-	DC Comics	676	Trickster	Male	blue	Human	Blond	183.0	-	81.0
Image Comics	426	Man of Miracles	-	blue	God / Eternal	Silver	-99.0	-	-99.0
Marvel Comics	33	Anti-Venom	Male	blue	Symbiote	Blond	229.0	-	358.0
Marvel Comics	110	Blackwulf	Male	red	Alien	White	188.0	-	88.0
Marvel Comics	692	Venompool	Male	-	Symbiote	-	226.0	-	-99.0

Subset rows from 'bad' to 'good'

In [34]:

heroes_srt.loc['bad':'good']

		sr	name	Gender	Eye color	Race	Hair color	Height	Skin color	Weight
Alignment	Publisher
bad	DC Comics	19	Amazo	Male	red	Android	-	257.0	-	173.0
DC Comics	31	Anti-Monitor	Male	yellow	God / Eternal	No Hair	61.0	-	-99.0
DC Comics	48	Atlas	Male	blue	God / Eternal	Brown	198.0	-	126.0
DC Comics	59	Bane	Male	-	Human	-	203.0	-	180.0
DC Comics	80	Big Barda	Female	blue	New God	Black	188.0	-	135.0
...	...	...	...	...	...	...	...	...	...	...
good	NaN	381	Katniss Everdeen	Female	-	Human	-	-99.0	-	-99.0
NaN	389	King Kong	Male	yellow	Animal	Black	30.5	-	NaN
NaN	393	Kool-Aid Man	Male	black	-	No Hair	-99.0	red	-99.0
NaN	542	Rambo	Male	brown	Human	Black	178.0	-	83.0
NaN	658	The Cape	Male	-	-	-	-99.0	-	-99.0

heroes_srt.loc[('good','DC Comics'):('neutral','Marvel Comics')]

Subset rows from 'good','DC Comics' to 'neutral','Marvel Comics'

 heroes_srt.loc[('good','DC Comics'):('neutral','Marvel Comics')]

Slicing series

Subsetting by row/column number

It is useful to subset rows with their row numbers. This is done with .iloc[], and like .loc[] it can take two arguments to let you subset by rows and columns.

To get the 5th row and 2nd column Value

In [42]:

heroes.iloc[4,1]

Out[42]:

'Abraxas'

To get First 5 Rows

In [43]:

heroes.iloc[:5]

	sr	name	Gender	Eye color	Race	Hair color	Height	Publisher	Skin color	Alignment	Weight
0	0	A-Bomb	Male	yellow	Human	No Hair	203.0	Marvel Comics	-	good	441.0
1	1	Abe Sapien	Male	blue	Icthyo Sapien	No Hair	191.0	Dark Horse Comics	blue	good	65.0
2	2	Abin Sur	Male	blue	Ungaran	No Hair	185.0	DC Comics	red	good	90.0
3	3	Abomination	Male	green	Human / Radiation	No Hair	203.0	Marvel Comics	-	bad	441.0
4	4	Abraxas	Male	blue	Cosmic Entity	Black	-99.0	Marvel Comics	-	bad	-99.0

To get All rows of Column 'name','Gender','Eye color':

In [44]:

heroes.iloc[:,1:4]

	name	Gender	Eye color
0	A-Bomb	Male	yellow
1	Abe Sapien	Male	blue
2	Abin Sur	Male	blue
3	Abomination	Male	green
4	Abraxas	Male	blue
...	...	...	...
729	Yellowjacket II	Female	blue
730	Ymir	Male	white
731	Yoda	Male	brown
732	Zatanna	Female	blue
733	Zoom	Male	red

To get 10 rows of Column 'name','Gender','Eye color':

In [45]:

heroes.iloc[:5,1:4]

	name	Gender	Eye color
0	A-Bomb	Male	yellow
1	Abe Sapien	Male	blue
2	Abin Sur	Male	blue
3	Abomination	Male	green
4	Abraxas	Male	blue