Using loc function of DataFrame

on
dataframe_using_loc

Skill - Selecting a subset of DataFrame using ‘loc’ function

Table of Contents

Skills Required

Please make sure to have all the skills mentioned above to understand and execute the code mentioned below. Go through the above skills if necessary for reference or revision

Pandas is a python library.
DataFrame is a data structure provided by the pandas library.

Please go through Pandas DataFrame Basics to learn the basics of pandas DataFrame.

In this post, we will learn how to select a subset of DataFrame using iloc function

The loc function of DataFrame can get a subset of DataFrame using the index values (for filtering rows) and column names (for filtering columns).

Suppose for a DataFrame df,

  • df.loc[[<index_list>], [<columns_list>]] returns a subset DataFrame
  • df.loc[[<True/False_list>], [<columns_list>]] returns a subset DataFrame
  • df.loc['<index>', '<column_name>'] returns a single value of the cell
  • df.loc['<index>', [<columns_list>]] will return the row data as a Series.
  • df.loc[[<index_list>], '<column_name>'] will return the column data as a Series.

Example: Getting values

import pandas as pd

# create a dataframe with column names and row indexes
df = pd.DataFrame([[2, 3], [5, 6], [8, 9]],
     index=['cobra', 'viper', 'sidewinder'],
     columns=['max_speed', 'shield'])
print(df)
'''
            max_speed  shield
cobra               2       3
viper               5       6
sidewinder          8       9
'''

# get a single row with index 'viper'
print(df.loc['viper'])
# we can see that it returns the row as a series
'''
max_speed    5
shield       6
Name: viper, dtype: int64
'''

# get two rows with index 'viper' and 'sidewinder'
print(df.loc[['viper', 'sidewinder']])
# we can see that it returns the the subset DataFrame
'''
            max_speed  shield
viper               5       6
sidewinder          8       9
'''

# get cell value with row index 'cobra' and column name 'shield'
print(df.loc['cobra', 'shield'])
# this should print 3

# get rows starting from 'cobra' till 'viper' and column 'max_speed'
print(df.loc['cobra':'viper', 'max_speed'])
# since we asked for only one column we got a series
'''
cobra    2
viper    5
Name: max_speed, dtype: int64
'''

# get rows with values in 'shield' column greater than 6
print(df.loc[df['shield'] > 6])
'''
            max_speed  shield
sidewinder          8       9
'''

# get rows with values in 'shield' column greater than 6, but return only 'max_speed' column
print(df.loc[df['shield'] > 6, ['max_speed']])
'''
            max_speed
sidewinder          8
'''

Example: Setting values

import pandas as pd

# create a dataframe with column names and row indexes
df = pd.DataFrame([[2, 3], [5, 6], [8, 9]],
     index=['cobra', 'viper', 'sidewinder'],
     columns=['max_speed', 'shield'])
print(df)
'''
            max_speed  shield
cobra               2       3
viper               5       6
sidewinder          8       9
'''

# set a single value for rows with indexes 'viper', 'sidewinder' 
# and column 'shield'
df.loc[['viper', 'sidewinder'], ['shield']] = 30
print(df)
'''
            max_speed  shield
cobra               2       3
viper               5      30
sidewinder          8      30
'''

# set a single value of all columns of a row with index 'cobra'
df.loc['cobra'] = 10
print(df)
'''
            max_speed  shield
cobra              10      10
viper               5      30
sidewinder          8      30
'''

# set a single value of all the rows with column name 'max_speed'
df.loc[:, 'max_speed'] = 30
print(df)
'''
            max_speed  shield
cobra              30      10
viper              30      30
sidewinder         30      30
'''

# set a single for all columns of rows that satisfy a condition
df.loc[df['shield'] > 25] = 0
print(df)
'''
            max_speed  shield
cobra              30      10
viper               0       0
sidewinder          0       0
'''

A similar function is iloc, but it uses row and column positions to get a subset of DataFrame

Video

Video for this post can be found here

Online Interpreter

Although we recommend to practice the above examples in Visual Studio Code, you can run these examples online at https://www.tutorialspoint.com/execute_python_online.php

References

Table of Contents

Comments

Post a Comment