Scribd
Upload
15 views

Pandas Practice 2

Uploaded by

yusuf01asif
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
15 views

Pandas Practice 2

Uploaded by

yusuf01asif
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

1.

​ ​Introduction

This assignment will help you to consolidate the concepts learnt in the session.

2.​ ​Problem Statement

import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

%matplotlib inline

df =
pd.read_csv('​https://raw.githubusercontent.com/jackiekazil/data-wrangling/master/data/chp3
/data-text.csv​')

df​.head(2)

df1 =
pd.read_csv('​https://raw.githubusercontent.com/kjam/data-wrangling-pycon/master/data/berl
in_weather_oldest.csv​')

df1​.head(2)

1. Get the Metadata from the above files.

Expected Output:
2. Get the row names from the above files.

Expected Output:

3. Change the column name from any of the above file.

Expected Output:

4. Change the column name from any of the above file and store the changes made
permanently.

Expected Output:

5. Change the names of multiple columns.

Expected Output:

6. Arrange values of a particular column in ascending order.


Expected Output:

7. Arrange multiple column values in ascending order.

Expected Output:

8. Make ​country​ as the first column of the dataframe.

Expected Output:
9. Get the column array using a variable

Expected Output:

10. Get the subset rows 11, 24, 37

Expected Output:

11. Get the subset rows excluding 5, 12, 23, and 56

Expected Output:

Load datasets from CSV

users =
pd.read_csv('​https://raw.githubusercontent.com/ben519/DataWrangling/master/Data/users.cs
v​')
sessions =
pd.read_csv('​https://raw.githubusercontent.com/ben519/DataWrangling/master/Data/session
s.csv​')

products =
pd.read_csv('​https://raw.githubusercontent.com/ben519/DataWrangling/master/Data/product
s.csv​')

transactions =
pd.read_csv('​https://raw.githubusercontent.com/ben519/DataWrangling/master/Data/transac
tions.csv​')

users.head()

sessions.head()

transactions.head()

12. Join users to transactions, keeping all rows from transactions and only matching rows from
users (left join)

Expected Output:
13. Which transactions have a UserID not in users?

Expected Output:

14. Join users to transactions, keeping only rows from transactions and users that match via
UserID (inner join)

Expected Output:

15. Join users to transactions, displaying all matching rows AND all non-matching rows (full
outer join)

Expected Output:
16. Determine which sessions occurred on the same day each user registered

Expected Output:

17. Build a dataset with every possible (UserID, ProductID) pair (cross join)

Expected Output:
18. Determine how much quantity of each product was purchased by each user

Expected Output:
19. For each user, get each possible pair of pair transactions (TransactionID1, TransacationID2)

Expected Output:
20. Join each user to his/her first occuring transaction in the transactions table

Expected Output:

21. Test to see if we can drop columns

Code with Output :

my_columns = list(data.columns)

my_columns

['UserID',
'User',
'Gender',
'Registered',
'Cancelled',
'TransactionID',
'TransactionDate',
'ProductID',
'Quantity']

list(data.dropna(thresh=int(data.shape[0] * .9), axis=1).columns) #set threshold to drop NAs

['UserID', 'User', 'Gender', 'Registered']

missing_info = list(data.columns[data.isnull().any()])

missing_info

['Cancelled', 'TransactionID', 'TransactionDate', 'ProductID', 'Quantity']

//for col in missing_info:

​ num_missing = data[data[col].isnull() == True].shape[0]

print('number missing for column {}: {}'.format(col, num_missing))

Output: Count of missing data

number missing for column Cancelled: 3


number missing for column TransactionID: 2
number missing for column TransactionDate: 2
number missing for column ProductID: 2
number missing for column Quantity: 2

//​for col in missing_info:

num_missing = data[data[col].isnull() == True].shape[0]

print('number missing for column {}: {}'.format(col, num_missing)) #count of missing data

for col in missing_info:

percent_missing = data[data[col].isnull() == True].shape[0] / data.shape[0]

print('percent missing for column {}: {}'.format(

col, percent_missing))

Output of percentage missing data

You might also like