How to Count Unique Values per Groups with Pandas?

Counting Unique Values per Groups with Pandas

When working with tabular data, it often becomes necessary to count the unique occurrences of values within specific groups. To achieve this in Python using the Pandas library, we can utilize the groupby() and nunique() methods.

Problem Explanation:

To illustrate the problem, consider the following dataset:

ID	domain
123	vk.com
123	vk.com
123	twitter.com
456	vk.com'
456	facebook.com
456	vk.com
456	google.com
789	twitter.com
789	vk.com

The task at hand is to count the unique ID values within each domain.

Solution:

To count unique values per group, we can use the following code:

<code class="python">df = df.groupby('domain')['ID'].nunique()</code>

The groupby() method groups the data by the domain column, while the nunique() method counts the unique occurrences of ID within each group. The output is a Series with the domain names as index and the corresponding unique counts as values.

domain
vk.com        3
twitter.com   2
facebook.com  1
google.com    1

Additional Notes:

• If the domain column values contain single quotes ('), you can remove them before grouping using the str.strip("'") method.
• To retain the column name in the output, use the agg() method with the pd.Series.nunique function.

Example with String Manipulation:

<code class="python">df['clean_domain'] = df.domain.str.strip("'")
df = df.groupby('clean_domain')['ID'].nunique()</code>

Example with agg():

<code class="python">df = df.groupby(by='domain', as_index=False).agg({'ID': pd.Series.nunique})</code>

The above is the detailed content of How to Count Unique Values per Groups with Pandas?. For more information, please follow other related articles on the PHP Chinese website!