replace_cluster_id_by_labels()

replace_cluster_id_by_labels() is a helper function that lets you replace numeric cluster IDs in a DataFrame with custom labels and optionally rename the identifier and cluster columns.

This is especially useful after running clustering, when you want to give clusters meaningful names (e.g., "Group A", "Group B") instead of numeric IDs.

Function Usage

new_df = replace_cluster_id_by_labels(
    df,
    mapping={1: "A", 2: "B", 3: "C"},  # optional
    new_cluster_column_name="Cluster",  # optional
    new_id_column_name="Entity ID"      # optional
)

Entry Parameters

Parameter	Required	Type	Description
df	✓	DataFrame	Input DataFrame containing at least "Entity ID" and "Cluster" columns.
mapping	✗	dict	Dictionary mapping cluster IDs (keys) to labels (values). Unmapped clusters remain unchanged.
new_cluster_column_name	✗	str	New name for the cluster column. Default = "Cluster".
new_id_column_name	✗	str	New name for the entity ID column. Default = "Entity ID".

What It Does

• Checks that your DataFrame contains both "Entity ID" and "Cluster" columns.

• Validates that all keys in mapping exist in the cluster IDs of your DataFrame.

• Replaces cluster IDs with custom labels based on the provided mapping.

  • Unmapped clusters remain as their original numeric values.

• Renames the identifier and cluster columns if you provide new names.

• Returns the updated DataFrame.

Returns

pd.DataFrame: A new DataFrame with cluster IDs replaced by labels and updated column names.

Key Features

• Custom labels: turn numeric cluster IDs into interpretable names.
• Validation: warns you if a mapping key does not exist in your DataFrame.
• Flexible renaming: lets you rename "Entity ID" and "Cluster" columns to fit your workflow.
• Safe handling: unmapped clusters stay untouched instead of breaking the DataFrame.

Examples

1. Replace cluster IDs with labels

import pandas as pd

original_df = pd.DataFrame({
    "Entity ID": [1, 2, 3],
    "Cluster": [1, 2, 3]
})

mapping = {1: "A", 2: "B", 3: "C"}

new_df = replace_cluster_id_by_labels(original_df, mapping)
print(new_df)

Output:

   Entity ID Cluster
0          1       A
1          2       B
2          3       C

2. Replace IDs and rename columns

new_df = replace_cluster_id_by_labels(original_df,
                                      mapping={1: "A", 2: "B", 3: "C"},
                                      new_cluster_column_name="Group",
                                      new_id_column_name="Person")
print(new_df)

Output:

   Person Group
0       1     A
1       2     B
2       3     C

See Also

• Data Preprocessing Overview maps the preparation pipeline.
• SequenceData is the next step after preprocessing.

Authors

Code: Yuqi Liang

Documentation: Yuqi Liang

Edited by: Yuqi Liang