How to Create Categorical Variables in Pandas (With Examples)

---

You can use one of the following methods to create a in pandas:

Method 1: Create Categorical Variable from Scratch

df['cat_variable'] = ['A', 'B', 'C', 'D']

Method 2: Create Categorical Variable from Existing Numerical Variable

df['cat_variable'] = pd.cut(df['numeric_variable'],
                            bins=[0, 15, 25, float('Inf')],
                            labels=['Bad', 'OK', 'Good'])

The following examples show how to use each method in practice.

Example 1: Create Categorical Variable from Scratch

The following code shows how to create a pandas DataFrame with one categorical variable called team and one numerical variable called points:

import pandas as pd

#create DataFrame with one categorical variable and one numeric variable
df = pd.DataFrame({'team': ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H'],
                   'points': [12, 15, 19, 22, 24, 25, 26, 30]})

#view DataFrame
print(df)

  team  points
0    A      12
1    B      15
2    C      19
3    D      22
4    E      24
5    F      25
6    G      26
7    H      30

#view data type of each column in DataFrame
print(df.dtypes)

team      object
points     int64
dtype: object

By using df.dtypes, we can  see the in the DataFrame.

We can see:

• The variable team is an object.
• The variable points is an integer.

In Python, an object is equivalent to a character or “categorical” variable. Thus, the team variable is a categorical variable.

Example 2: Create Categorical Variable from Existing Numerical Variable

The following code shows how to create a categorical variable called status from the existing numerical variable called points in the DataFrame:

import pandas as pd

#create DataFrame with one categorical variable and one numeric variable
df = pd.DataFrame({'team': ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H'],
                   'points': [12, 15, 19, 22, 24, 25, 26, 30]})

#create categorical variable 'status' based on existing numerical 'points' variable
df['status'] = pd.cut(df['points'],
                      bins=[0, 15, 25, float('Inf')],
                      labels=['Bad', 'OK', 'Good'])

#view updated DataFrame
print(df)

  team  points status
0    A      12    Bad
1    B      15    Bad
2    C      19     OK
3    D      22     OK
4    E      24     OK
5    F      25     OK
6    G      26   Good
7    H      30   Good

Using the cut() function, we created a new categorical variable called status that takes the following values:

• ‘Bad‘ if the value in the points column is less than or equal to 15.
• Else, ‘OK‘ if the value in the points column is less than or equal to 25.
• Else, ‘Good‘.

Note that when using the cut() function, the number of labels must be one less than the number of bins.

In our example, we used four values for bins to define the bin edges and three values for labels to specify the labels to use for the categorical variable.