You can use the nunique() function to count the number of unique values in a pandas DataFrame.

This function uses the following basic syntax:

#count unique values in each column
df.nunique()

#count unique values in each row
df.nunique(axis=1)

The following examples show how to use this function in practice with the following pandas DataFrame:

import pandas as pd

#create DataFrame
df = pd.DataFrame({'team': ['A', 'A', 'A', 'A', 'B', 'B', 'B', 'B'],
                   'points': [8, 8, 13, 13, 22, 22, 25, 29],
                   'assists': [5, 8, 7, 9, 12, 9, 9, 4],
                   'rebounds': [11, 8, 11, 6, 6, 5, 9, 12]})

#view DataFrame
df

	team	points	assists	rebounds
0	A	8	5	11
1	A	8	8	8
2	A	13	7	11
3	A	13	9	6
4	B	22	12	6
5	B	22	9	5
6	B	25	9	9
7	B	29	4	12

Example 1: Count Unique Values in Each Column

The following code shows how to count the number of unique values in each column of a DataFrame:

#count unique values in each column
df.nunique()

team        2
points      5
assists     5
rebounds    6
dtype: int64

From the output we can see:

• The ‘team’ column has 2 unique values
• The ‘points’ column has 5 unique values
• The ‘assists’ column has 5 unique values
• The ‘rebounds’ column has 6 unique values

Example 2: Count Unique Values in Each Row

The following code shows how to count the number of unique values in each row of a DataFrame:

#count unique values in each row
df.nunique(axis=1)

0    4
1    2
2    4
3    4
4    4
5    4
6    3
7    4
dtype: int64

From the output we can see:

• The first row has 4 unique values
• The second row has 2 unique values
• The third row has 4 unique values

And so on.

Example 3: Count Unique Values by Group

The following code shows how to count the number of unique values by group in a DataFrame:

#count unique 'points' values, grouped by team
df.groupby('team')['points'].nunique()

team
A    2
B    3
Name: points, dtype: int64

From the output we can see:

• Team ‘A’ has 2 unique ‘points’ values
• Team ‘B’ has 3 unique ‘points’ values

Example 4: Count Unique Values with a Condition

You can also combine nunique() with filtering to count unique values that meet specific conditions. For instance, let’s count the number of unique values where points are greater than 10:

#count unique values where points > 10
df[df['points'] > 10].nunique()

team        2
points      4
assists     4
rebounds    5
dtype: int64

From the output we can see:

• There are 2 unique teams with points > 10
• There are 4 unique point values > 10
• There are 4 unique assist values for rows with points > 10
• There are 5 unique rebound values for rows with points > 10

Example 5: Count Unique Values Using value_counts()

While nunique() gives you the count of unique values, value_counts() provides both the unique values and their frequencies. Let’s see how this works with our dataset:

#count occurrences of each unique value in the 'points' column
df['points'].value_counts()

8     2
13    2
22    2
25    1
29    1
Name: points, dtype: int64

From the output we can see:

• The value 8 appears 2 times
• The value 13 appears 2 times
• The value 22 appears 2 times
• The value 25 appears 1 time
• The value 29 appears 1 time

We can confirm that there are 5 unique values in the ‘points’ column, which matches our earlier nunique() result.

Wrapping Up

The nunique() function in pandas provides a straightforward way to count unique values in your data. This can be useful for:

• Understanding data variety in each column
• Checking for duplicate values
• Analyzing data distribution by groups
• Determining the cardinality of categorical variables

By using nunique() along with other pandas functions like groupby() and value_counts(), you can gain deeper insights into your data’s composition and structure.

Additional Resources

The following tutorials explain how to perform other common operations in pandas:

How to Count Observations by Group in Pandas
How to Count Missing Values in Pandas
How to Use Pandas value_counts() Function