Split dataframe in Pandas based on values in multiple columns

Last Updated : 31 Jul, 2023

In this article, we are going to see how to divide a dataframe by various methods and based on various parameters using Python. To divide a dataframe into two or more separate dataframes based on the values present in the column we first create a data frame.

Creating a DataFrame for demonestration

Python3

# importing pandas as pd import pandas as pd   # dictionary of lists dict = {'First_Name': ["Aparna", "Pankaj", "Sudhir",                         "Geeku", "Anuj", "Aman",                        "Madhav", "Raj", "Shruti"],         'Last_Name': ["Pandey", "Gupta", "Mishra",                        "Chopra", "Mishra", "Verma",                        "Sen", "Roy", "Agarwal"],         'Email_ID': ["[email protected]", "[email protected]",                      "[email protected]", "[email protected]",                      "[email protected]", "[email protected]",                      "[email protected]", "[email protected]",                      "[email protected]"],         'Degree': ["MBA", "BCA", "M.Tech", "MBA", "B.Sc",                    "B.Tech", "B.Tech", "MBA", "M.Tech"],         'Score': [90, 40, 75, 98, 94, 90, 80, 90, 95]}  # creating dataframe df = pd.DataFrame(dict)  print(df)

Output:

Split dataframe based on values By Boolean Indexing

We can create multiple dataframes from a given dataframe based on a certain column value by using the boolean indexing method and by mentioning the required criteria.

Example 1: Creating a dataframe for the students with Score >= 80

Python3

# creating a new dataframe by applying the required # conditions in [] df1 = df[df['Score'] >= 80] print(df1)

Output:

Example 2: Creating a dataframe for the students with Last_Name as Mishra

Python3

# Creating on the basis of Last_Name dfname = df[df['Last_Name'] == 'Mishra'] print(dfname)

Output:

We can do the same for other columns as well by putting the appropriate condition

Split dataframe based on values Boolean Indexing with mask variable

We create a mask variable for the condition of the column in the previous method

Example 1: To get dataframe of students with Degree as MBA

Python3

# creating the mask variable with appropriate # condition mask_var = df['Degree'] =='MBA'  # creating a dataframe df1_mask = df[mask_var]  print(df1_mask)

Output :

Example 2: To get a dataframe for the rest of the students

To get the rest of the values in a dataframe we can simply invert the mask variable by adding a ~(tilde) after it.

Python3

# creating dataframe with inverted mask variable df2_mask = df[~mask_var] print(df2_mask)

Output :

Split dataframe based on values Using groupby() function

Using groupby() we can group the rows using a specific column value and then display it as a separate dataframe.

Example 1: Group all Students according to their Degree and display as required

Python3

# Creating an object using groupby grouped = df.groupby('Degree')  # the return type of the object 'grouped' is  # pandas.core.groupby.generic.DataFrameGroupBy.  # Creating a dataframe from the object using get_group(). # dataframe of students with Degree as MBA. df_grouped = grouped.get_group('MBA')  print(df_grouped)

Output: dataframe of students with Degree as MBA

Example 2: Group all Students according to their Score and display as required

Python3

# Creating another object using groupby grouped2 = df.groupby('Score')  # the return type of the object 'grouped2' is  # pandas.core.groupby.generic.DataFrameGroupBy.  # Creating a dataframe from the object  # using get_group() dataframe of students # with Score = 90 df_grouped2 = grouped2.get_group(90)  print(df_grouped2)