6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.

ipynb - Colaboratory

#python pandas
#used to work with heterogenous (tabular) data
#facilitate faster and easier data processing

Double-click (or enter) to edit

import pandas as pd

# frequently used libraries

from pandas import Series, DataFrame

## Data structures: Series & Dataframes

## Series
#A sequence of values ( similar to 1 D array) with explicit custom index to access the values
#like a fixed length dict. a KIND mapping of index values to data values
# Created using Series function
series1 = pd.Series([22,33,44,55]) # default index: 0 to n-1

series1

0 22
1 33
2 44
3 55
dtype: int64

# Series properties (attributes)

# 1. 'values'
#it is an array object
series1.values

array([22, 33, 44, 55])

# 2. index
# it is like range
series1.index

RangeIndex(start=0, stop=4, step=1)

#custom index: as a list of labels

series2 = pd.Series([22,33,44,55], index=['a','b','c','d'])

series2

a 22
b 33
c 44
d 55
dtype: int64

# check presence or absence of labels in a sereis

'c' in series2

True

'e' in series2

False

series2.values

array([22, 33, 44, 55])

series2.index # to run code press shift + enter

Index(['a', 'b', 'c', 'd'], dtype='object')

# accessing values using labeled indices

# series2(['b', 'a']) #error as incorrect brakets

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 1/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory
series2[['b','a']]

b 33
a 22
dtype: int64

#index-value link is not disturbed by execution of various operations.

# automatic alignment of index labels with corresponding values and results just like in 'R' platform
# filtering
series2[series2>0]

a 22
b 33
c 44
d 55
dtype: int64

series2[series2>34]

c 44
d 55
dtype: int64

# scalar multiplication
series2*9 # Broadcasting is done here.

a 198
b 297
c 396
d 495
dtype: int64

#math operation
import numpy as np
np.exp(series2) # exponential of series values

a 3.584913e+09
b 2.146436e+14
c 1.285160e+19
d 7.694785e+23
dtype: float64

# Converting Python dict object into a series object as mapping is possible

# PIN_code = {'Roorkee'= 247667,'Dehradun'= 248001, 'Haridwar' = 249401}
# above invalid syntax = should be replaced with :
PIN_code = {'Roorkee': 247667, 'Dehradun': 248001, 'Haridwar': 249401}

series3 = pd.Series(PIN_code)

series3

Roorkee 247667
Dehradun 248001
Haridwar 249401
dtype: int64

series3.index

Index(['Roorkee', 'Dehradun', 'Haridwar'], dtype='object')

# METHOD 1 of changing INDEX

# changing index using index argument
# e.g., order of indices
cities = ['Dehradun', 'Roorkee', 'Haridwar', 'Tehri garhwal']

series4 = pd.Series(PIN_code, index=cities)

series4

Dehradun 248001.0
Roorkee 247667.0
Haridwar 249401.0
Tehri garhwal NaN
dtype: float64

# detect missing data

# using isnul & not null function
pd.isnull(series4)

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 2/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

Dehradun False
Roorkee False
Haridwar False
Tehri garhwal True
dtype: bool

pd.notnull(series4)

Dehradun True
Roorkee True
Haridwar True
Tehri garhwal False
dtype: bool

# using isnull and notnull methods

series4.isnull()

Dehradun False
Roorkee False
Haridwar False
Tehri garhwal True
dtype: bool

series4.notnull()

Dehradun True
Roorkee True
Haridwar True
Tehri garhwal False
dtype: bool

# alignment feature
series3 + series4 # pincode added.

Dehradun 496002.0
Haridwar 498802.0
Roorkee 495334.0
Tehri garhwal NaN
dtype: float64

# name attributes of series object and its index

series4.name = 'PIN_code' #see last line of below execution result

series4.index.name = 'city' # to specify name of index

series4

city
Dehradun 248001.0
Roorkee 247667.0
Haridwar 249401.0
Tehri garhwal NaN
Name: PIN_code, dtype: float64

# METHOD 2 of changing INDEX

# changing index using index attribute
# e.g index values
series1

0 22
1 33
2 44
3 55
dtype: int64

series1.index = [1,2,3,4]

series1

1 22
2 33
3 44
4 55
dtype: int64

#DATAFRAME Represents Physically 2-D data in a tabular form row column

#It is like a dict of series elements having a common (row) index

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 3/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

#it has both a row & a column index

##row index is similar to series index (default 0 to n-1)

#column index is variables name (by default sorted alphabetically)

#columns can be variables of different value types (numeric, string, boolean,etc)

#created using Dataframe function

# ltes take year wise metro population data

# Dataframe is like a dict with : common key with multiple elements
dict1 = {'metro':['Delhi','Delhi','Delhi','Mumbai', 'Mumbai', 'Mumbai'],
'year': [2011,2012,2013,2011,2012,2013],
'popcr':[1.67,1.72,1.77,2.07,2.11,2.15]}

#possible error
# df = pd.dataframe(dict1)
# df = pd.Dataframe(dict1)

df = pd.DataFrame(dict1)

df

metro year popcr

0 Delhi 2011 1.67

1 Delhi 2012 1.72

2 Delhi 2013 1.77

3 Mumbai 2011 2.07

4 Mumbai 2012 2.11

5 Mumbai 2013 2.15

#first five row

df.head()

metro year popcr

0 Delhi 2011 1.67

1 Delhi 2012 1.72

2 Delhi 2013 1.77

3 Mumbai 2011 2.07

4 Mumbai 2012 2.11

# changing the column index using columns argument

pd.DataFrame(dict1, columns =['year','metro','popcr'])

year metro popcr

0 2011 Delhi 1.67

1 2012 Delhi 1.72

2 2013 Delhi 1.77

3 2011 Mumbai 2.07

4 2012 Mumbai 2.11

5 2013 Mumbai 2.15

# chnaging the row index using index argument

df2 = pd.DataFrame(dict1, columns =['year','metro','popcr','area'],index=list(range(1,7)))
df2

# row and columnindex both changed.

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 4/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

year metro popcr area

1 2011 Delhi 1.67 NaN

2 2012 Delhi 1.72 NaN

3 2013 Delhi 1.77 NaN

4 2011 Mumbai 2.07 NaN

# column index
df2.columns
5 2012 Mumbai 2.11 NaN

Index(['year',
6 2013 Mumbai 'metro',
2.15 'popcr',
NaN 'area'], dtype='object')

# accessing column using dict key like notation

df2['popcr']

1 1.67
2 1.72
3 1.77
4 2.07
5 2.11
6 2.15
Name: popcr, dtype: float64

# accessing column using attribute like notation

df2.year

1 2011
2 2012
3 2013
4 2011
5 2012
6 2013
Name: year, dtype: int64

# selecting particular row indices along a column

df2['popcr'][1:4]

2 1.72
3 1.77
4 2.07
Name: popcr, dtype: float64

df2['popcr'][:5] # only stopping point

1 1.67
2 1.72
3 1.77
4 2.07
5 2.11
Name: popcr, dtype: float64

df2['popcr'][:-3]

1 1.67
2 1.72
3 1.77
Name: popcr, dtype: float64

df2['popcr'][:] # no start stop, all observations shown

1 1.67
2 1.72
3 1.77
4 2.07
5 2.11
6 2.15
Name: popcr, dtype: float64

# accessing row using 'loc' attribute (locate)

df2.loc[:]

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 5/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

year metro popcr area

1 2011 Delhi 1.67 NaN

2 2012 Delhi 1.72 NaN

df2.loc[1]
3 2013 Delhi 1.77 NaN
year 2011
4 2011 Delhi
metro Mumbai 2.07 NaN
popcr 1.67
5 2012 Mumbai 2.11 NaN
area NaN
Name: 1, dtype:
6 2013 Mumbai object
2.15 NaN

## Data processing & transformation

# value assignment
# e.g area column
df2['area']=6000

df2

year metro popcr area

1 2011 Delhi 1.67 6000

2 2012 Delhi 1.72 6000

3 2013 Delhi 1.77 6000

4 2011 Mumbai 2.07 6000

5 2012 Mumbai 2.11 6000

6 2013 Mumbai 2.15 6000

# populating values
#e.g area column
df2['area'] = np.arange(6000,6600,100) # (start, stop, step)
df2

year metro popcr area

1 2011 Delhi 1.67 6000

2 2012 Delhi 1.72 6100

3 2013 Delhi 1.77 6200

4 2011 Mumbai 2.07 6300

5 2012 Mumbai 2.11 6400

6 2013 Mumbai 2.15 6500

# using series to populate vaules

# unmtched indices would be treated as missing values
val = pd.Series([1.7,1.9,2.1], index=[2,4,6])
df2['popcr'] = val
df2

year metro popcr area

1 2011 Delhi NaN 6000

2 2012 Delhi 1.7 6100

3 2013 Delhi NaN 6200

4 2011 Mumbai 1.9 6300

5 2012 Mumbai NaN 6400

6 2013 Mumbai 2.1 6500

# assigning values to a non existing column will also create that column
df2['northern'] = (df2.metro == 'Delhi') # kind of boolean column

df2

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 6/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

year metro popcr area northern

1 2011 Delhi NaN 6000 True

2 2012 Delhi 1.7 6100 True

3 2013 Delhi NaN 6200 True

4 2011 Mumbai 1.9 6300 False

# removing a column
5 2012 Mumbai NaN 6400 False
del df2['northern']
6 2013 Mumbai 2.1 6500 False

df2.columns

Index(['year', 'metro', 'popcr', 'area'], dtype='object')

#filling values when reindexing

# using ffill method
series10 = pd.Series(['blue','purple','yellow'], index=[0,2,4])

## more on 'index objects' in series and DataFrame

# immutable array-like objects
# Also store metadata other than axis labels (index values)

# sequences of labels used while creating a series or dataframe are internally converted into
series5 = pd.Series(range(3), index=['a','b','c']) # index square braket

series5.index

Index(['a', 'b', 'c'], dtype='object')

#subsetting
ind1 = series5.index
ind1[1:] #metion starting point

Index(['b', 'c'], dtype='object')

ind1[0:2]

Index(['a', 'b'], dtype='object')

#immutability
ind1[1] = 'd'
# output is error : type error
# index does not support mutable operation

---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-126-65fbb217ef82> in <cell line: 2>()
1 #immutability
----> 2 ind1[1] = 'd'
3 # output is error : type error
4 # index does not support mutable operation

/usr/local/lib/python3.10/dist-packages/pandas/core/indexes/base.py in
__setitem__(self, key, value)
5300 @final
5301 def __setitem__(self, key, value):
-> 5302 raise TypeError("Index does not support mutable operations")
5303
5304 def __getitem__(self, key):

TypeError: Index does not support mutable operations

SEARCH STACK OVERFLOW

# creating index using Index function

ind2 = pd.Index(np.arange(3))

ind2

Int64Index([0, 1, 2], dtype='int64')

series6 = pd.Series([1.5,-2.5,0], index=ind2)

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 7/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory
series6

0 1.5
1 -2.5
2 0.0
dtype: float64

series6.index is ind2

True

# Index objects can have duplicate elements

ind3 = pd.Index(['beta','alpha','beta','sigma'])

ind3

Index(['beta', 'alpha', 'beta', 'sigma'], dtype='object')

series7 = pd.Series([1.1,1.2,1.3,1.4], index=ind3)

series7

beta 1.1
alpha 1.2
beta 1.3
sigma 1.4
dtype: float64

# two rows with same index beta, lets see

series7['beta'] # shows all rows with label beta

beta 1.1
beta 1.3
dtype: float64

series10

0 blue
2 purple
4 yellow
dtype: object

series11 = series10.reindex(range(6), method = 'ffill')

series11

0 blue
1 blue
2 purple
3 purple
4 yellow
5 yellow
dtype: object

# reindexing with dataframe

df4 = pd.DataFrame((np.arange(9).reshape(3,3)),index=['a','c','d'], columns=['Delhi','Mumbai','Bangalore'])

df4

Delhi Mumbai Bangalore

a 0 1 2

c 3 4 5

d 6 7 8

#reindexing rows index

df5 = df4.reindex(['a','b','c','d'])

df5

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 8/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

Delhi Mumbai Bangalore

a 0.0 1.0 2.0

b NaN NaN NaN

c 3.0 4.0 5.0

#reindexing columns index
#usingdcolumns
6.0 argument
7.0 8.0
df5.reindex(columns=['Banglore','Delhi','Mumbai'])

Banglore Delhi Mumbai

a NaN 0.0 1.0

b NaN NaN NaN

c NaN 3.0 4.0

d NaN 6.0 7.0

# droping cases or rows

df5.drop('b') # df5.drop(['a','c']) dropping more rows by passing list argument

Delhi Mumbai Bangalore

a 0.0 1.0 2.0

c 3.0 4.0 5.0

d 6.0 7.0 8.0

#Dropping columns or variables

df5.drop('Bangalore', axis=1) # we hv to specify axis, as to confirm that column is to be dropped

Delhi Mumbai

a 0.0 1.0

b NaN NaN

c 3.0 4.0

d 6.0 7.0

# df5.drop('Delhi','Mumbai'], axis=1) # dropping multiple columns

# using in-place version of drop method

# using implace argument
df5

Delhi Mumbai Bangalore

a 0.0 1.0 2.0

b NaN NaN NaN

c 3.0 4.0 5.0

d 6.0 7.0 8.0

df5.drop('b', inplace = True) #

df5

Delhi Mumbai Bangalore

a 0.0 1.0 2.0

c 3.0 4.0 5.0

d 6.0 7.0 8.0

# Ex. access, selection, and filtering

# Working with a series
series11

0 blue
1 blue
2 purple
3 purple
4 yellow

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 9/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory
5 yellow
dtype: object

series7_1 = pd.DataFrame([1.1,1.3,1.3,1.3],index= ['beta','alpha','beta','sigma'])

series7_1

beta 1.1

alpha 1.3

beta 1.3

sigma 1.3

# accessing one row

series11[3] # square brakets

'purple'

series7_1['sigma'] # square braket with single quotes

# 1.3

---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
<ipython-input-153-8bdd86d373a9> in <cell line: 1>()
----> 1 series7_1['sigma'] # square braket with single quotes
2 # 1.3

1 frames
/usr/local/lib/python3.10/dist-packages/pandas/core/indexes/range.py in
get_loc(self, key, method, tolerance)
393 raise KeyError(key) from err
394 self._check_indexing_error(key)
--> 395 raise KeyError(key)
396 return super().get_loc(key, method=method, tolerance=tolerance)
397

KeyError: 'sigma'

SEARCH STACK OVERFLOW

# accessing a range of rows

series11[3:5]

3 purple
4 yellow
dtype: object

# selecting few rows

series7_1[['sigma','alpha']] #in index use square brakets always

---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
<ipython-input-156-b23e739c9d06> in <cell line: 2>()
1 # selecting few rows
----> 2 series7_1[['sigma','alpha']] #in index use square brakets always

2 frames
/usr/local/lib/python3.10/dist-packages/pandas/core/indexes/base.py in
_raise_if_missing(self, key, indexer, axis_name)
6128 if use_interval_msg:
6129 key = list(key)
-> 6130 raise KeyError(f"None of [{key}] are in the [{axis_name}]")
6131
6132 not_found = list(ensure_index(key)[missing_mask.nonzero()
[0]].unique())

KeyError: "None of [Index(['sigma', 'alpha'], dtype='object')] are in the [columns]"

# Using Earlier created series 7

# selecting few rows
series7[['sigma','alpha']]

sigma 1.4
alpha 1.2
dtype: float64

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 10/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory
# filtering rows based on some logic
series7[series7<1.3]

beta 1.1
alpha 1.2
dtype: float64

# slicing

# Note : end-point is inclusive in series and dataframe

# series7['alpha','sigma'] shows error
series7['alpha':'sigma']

alpha 1.2
beta 1.3
sigma 1.4
dtype: float64

series7['alpha':'sigma'] = 1.3 # asssigning 1.3

series7 # values assigned

beta 1.1
alpha 1.3
beta 1.3
sigma 1.3
dtype: float64

series7['beta']

beta 1.1
beta 1.3
dtype: float64

## interaction mechanism with the data in a series or dataframe

# Ex. reindexing
# using redindex method
series8 = pd.Series([4.5, 7.2, -5.3, 3.6], index = ['d', 'b', 'a', 'c'])

#pd.Series([],index= [])

series8

d 4.5
b 7.2
a -5.3
c 3.6
dtype: float64

series9 = series8.reindex(['a','b','c','d','e'])

series9

a -5.3
b 7.2
c 3.6
d 4.5
e NaN
dtype: float64

# filling values whe reindexing

# using ffill method

series10 = pd.Series(['blue','purple','yellow'],index = [0,2,4])

series10

0 blue
2 purple
4 yellow
dtype: object

series11 = series10.reindex(range(6), method = 'ffill')

series11

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 11/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

0 blue
1 blue
2 purple
3 purple
4 yellow
5 yellow
dtype: object

# reindexing with dataframe

df4 = pd.DataFrame(np.arange(9).reshape((3,3)), index=['a','c','d'], columns = ['Delhi','Mumbai','Banglore'])

# Error chnaces
#1. index=['a','b',c','d']
# brakets

df4

Delhi Mumbai Banglore

a 0 1 2

c 3 4 5

d 6 7 8

# reindexing rows index

df5 = df4.reindex(['a','b','c','d'])

df5

Delhi Mumbai Banglore

a 0.0 1.0 2.0

b NaN NaN NaN

c 3.0 4.0 5.0

d 6.0 7.0 8.0

# reindexing column index, using column argument

df5.reindex(columns = ['Banglore','Delhi','Mumbai'])

Banglore Delhi Mumbai

a 2.0 0.0 1.0

b NaN NaN NaN

c 5.0 3.0 4.0

d 8.0 6.0 7.0

# reindexing using loc attribute

df4.loc[['a','b','c','d'],['Banglore','Delhi','Mumbai']]

# previously its working but not in recent updates

A function that takes the passenger name as input, checks which booking class the passenger belongs to, and returns the average delay time.

# Accept passenger name input from user

passenger_name = input("Enter passenger name: ")

booking_class = df.loc[df['Name'] == passenger_name , 'Booking Class'].iloc[0]

print("Booking class for passenger", passenger_name, ":", booking_class)
delay_min = df.loc[df['Name'] == passenger_name , 'Delay Minutes'].iloc[0]
print("Delay minutes for passenger", passenger_name, ":", delay_min)

Enter passenger name: Amber Mcclure Booking class for passenger Amber Mcclure : First Delay minutes for passenger Amber Mcclure : 8

# Ex. dropping cases or rows

## using drop method

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 12/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

df5

Delhi Mumbai Banglore

a 0.0 1.0 2.0

b NaN NaN NaN

c 3.0 4.0 5.0

d 6.0 7.0 8.0

df5.drop('b')

Delhi Mumbai Banglore

a 0.0 1.0 2.0

c 3.0 4.0 5.0

d 6.0 7.0 8.0

# df5.drop('a','c') error
# have to pass list
df5.drop(['a','c'])

Delhi Mumbai Banglore

b NaN NaN NaN

d 6.0 7.0 8.0

df5.drop('Banglore', axis=1)

Delhi Mumbai

a 0.0 1.0

b NaN NaN

c 3.0 4.0

d 6.0 7.0

# dropping multiple columns

# multiple columns in square bracket followed by ,

df5.drop(['Delhi','Mumbai'], axis=1)

Banglore

a 2.0

b NaN

c 5.0

d 8.0

# using in-place version of drop method

#using inplace argument

df5

Delhi Mumbai Banglore

a 0.0 1.0 2.0

b NaN NaN NaN

c 3.0 4.0 5.0

d 6.0 7.0 8.0

df5.drop('b', inplace = True)

df5

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 13/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

Delhi Mumbai Banglore

a 0.0 1.0 2.0

c 3.0 4.0 5.0

d 6.0 7.0 8.0

# access, selection, filtering

# working with a series

series11

0 blue
1 blue
2 purple
3 purple
4 yellow
5 yellow
dtype: object

series7

beta 1.1
alpha 1.3
beta 1.3
sigma 1.3
dtype: float64

# accessing one row

series11[3] #integer index 3

'purple'

series7['sigma'] # string index

1.3

# accessing a range of row

series11[3:5]

3 purple
4 yellow
dtype: object

# selecting few row

series7[['sigma','alpha']]

sigma 1.3
alpha 1.3
dtype: float64

series11[[1,5]] # selecting row index 1 & 5

1 blue
5 yellow
dtype: object

# filtering rows based on some logic

series7[series7<1.3]

beta 1.1
dtype: float64

# slicing
# Note: end point inclusive on series and dataframe

series7['alpha':'sigma']

#series7['alpha','sigma'] not work

alpha 1.3
beta 1.3
sigma 1.3
dtype: float64

series7['alpha':'sigma'] = 1.3

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 14/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

series7

beta 1.1
alpha 1.3
beta 1.3
sigma 1.3
dtype: float64

#### New Topic Starts

# working with DataFrame
df5

Delhi Mumbai Bangalore

a 0.0 1.0 2.0

c 3.0 4.0 5.0

d 6.0 7.0 8.0

# accessing columns
df5['Delhi']

a 0.0
c 3.0
d 6.0
Name: Delhi, dtype: float64

df5[['Delhi','Bangalore']]

Delhi Bangalore

a 0.0 2.0

c 3.0 5.0

d 6.0 8.0

# slicing rows
df5[:2] # starting point not specified

Delhi Mumbai Bangalore

a 0.0 1.0 2.0

c 3.0 4.0 5.0

df5[1:2]

Delhi Mumbai Bangalore

c 3.0 4.0 5.0

# filtering rows
df5[df5['Bangalore']>5]

Delhi Mumbai Bangalore

d 6.0 7.0 8.0

# working with a boolean dataframe

df5 < 5

Delhi Mumbai Bangalore

a True True True

c True True False

d False False False

df5[df5<5] = 0

df5

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 15/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

Delhi Mumbai Bangalore

a 0.0 0.0 0.0

c 0.0 0.0 5.0

## indexing
d operators
6.0 7.0 loc and iloc
8.0 , gives precise indexing
# loc is used with axis labels
# iloc is used with integer values corresponding to axis labels
df5

Delhi Mumbai Bangalore

a 0.0 0.0 0.0

c 0.0 0.0 5.0

d 6.0 7.0 8.0

#syntax: dataframe.loc[row,column] square brakets

# accessing sigle elemnt
df5.loc['a','Delhi']

0.0

# df5.loc['a','Delhi','Bangalore'] Gives indexing error

df5.loc['a',['Delhi','Banglore']]

Delhi 0.0
Banglore 2.0
Name: a, dtype: float64

# df5.loc['a','b'],['Delhi','Bangalore']
# not allowed now , earlier shows results.

df5.iloc[0,0] #[row, column]

0.0

df5.iloc[0,[0,1]] # selecting two columns

Delhi 0.0
Mumbai 0.0
Name: a, dtype: float64

df5.iloc[[0,1],[0,1]] # 2 row , 2 columns

Delhi Mumbai

a 0.0 0.0

c 0.0 0.0

df5.iloc[1] # second row

Delhi 0.0
Mumbai 0.0
Bangalore 5.0
Name: c, dtype: float64

df5.iloc[[1,2],[2,0,1]] # 2 row , 3 columns

Bangalore Delhi Mumbai

c 5.0 0.0 0.0

d 8.0 6.0 7.0

# slicing with loc and iloc

df5.loc[:'c', 'Delhi']

a 0.0
c 0.0
Name: Delhi, dtype: float64

df5.loc['c':'d', 'Delhi']

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 16/17
6/25/23, 10:02 PM Copy of Copy of IIT R DIXIT DATAFRAMES.ipynb - Colaboratory

c 0.0
d 6.0
Name: Delhi, dtype: float64

df5.loc['c':'d', 'Delhi':'Bangalore']

Delhi Mumbai Bangalore

c 0.0 0.0 5.0

d 6.0 7.0 8.0

df5.iloc[:, 1:2]
# : shows all rows, all columns are to selected
# on column side we are selecting 1,2

Mumbai

a 0.0

c 0.0

d 7.0

df5.iloc[:,0:2][df5.Mumbai != 0]
# Filtering apart from slicing
# all rows, 0 to 2 columns are going to be selected
# only those columns are selected in which value is not equal to 0

Delhi Mumbai

d 6.0 7.0

# # more on integer indexing

series11

0 blue
1 blue
2 purple
3 purple
4 yellow
5 yellow
dtype: object

# fails due to inference ambiguity

series11[-3]

series7[-1] #doesnot fails

1.3

check 0s completed at 10:00 PM

https://colab.research.google.com/drive/1KY8qW02fPGwZI8S3tRjhoj8Z1alZWqCf#scrollTo=iDltQkrkPOd6&uniqifier=1&printMode=true 17/17