How to select the rows of a dataframe using the indices of another dataframe?

Last Updated : 26 Dec, 2020

Prerequisites:

Using Pandas module it is possible to select rows from a data frame using indices from another data frame. This article discusses that in detail. It is advised to implement all the codes in jupyter notebook for easy implementation.

Approach:

Import module
Create first data frame. In the example given below choice(), randint() and random() all belonging to random module are used to generate a data frame.

1) choice() – choice() is an inbuilt function in Python programming language that returns a random item from a list, tuple, or string.

Syntax: random.choice(sequence)

Parameters: Sequence is a mandatory parameter that can be a list, tuple, or string.

Returns: The choice() returns a random item.

2) randint()- This function is used to generate random numbers

Syntax : randint(start, end)

Parameters :

(start, end) : Both of them must be integer type values.

Returns :

A random integer in range [start, end] including the end points.

3) random()- Used to generate floating numbers between 0 and 1.

Create another data frame using the random() function and randomly selecting the rows of the first dataset.
Now we will use dataframe.loc[] function to select the row values of the first data frame using the indexes of the second data frame. Pandas DataFrame.loc[] attribute access a group of rows and columns by label(s) or a boolean array in the given DataFrame.

Syntax: DataFrame.loc

Parameter : None

Returns : Scalar, Series, DataFrame

Display selected rows

Implementation using the above concept is given below:

Program:

Python3

# Importing Required Libraries 
import pandas as pd 
import random 
  
# Creating data for main dataframe 
col1 = [random.randint(1, 9) for i in range(15)] 
col2 = [random.random() for i in range(15)] 
col3 = [random.choice(['Geeks', 'of', 'Computer', 'Science']) 
        for i in range(15)] 
col4 = [random.randint(1, 9) for i in range(15)] 
col5 = [random.randint(1, 9) for i in range(15)] 
  
# Defining Column name for main dataframe 
data_generated = { 
    'value1': col1, 
    'value2': col2, 
    'value3': col3, 
    'value4': col4, 
    'value5': col5 
} 
  
# Creating the datafram using DataFrame() function 
print("First data frame") 
dataframe = pd.DataFrame(data_generated) 
display(dataframe) 
  
# Creating a second dataframe that will be the subset of main dataframe 
print("Second data frame") 
dataframe_second = dataframe[['value1', 'value2', 'value3']].sample(n=4) 
display(dataframe_second) 
  
# Rows of a dataframe using the indices of another dataframe 
print("selecting rows of first datafame using second dataframe") 
display(dataframe.loc[dataframe_second.index]) 