Skip to content

frequency

  • Name: cognitivefactory.interactive_clustering.utils.frequency
  • Description: Utilities methods for frequency analysis.
  • Author: Erwan SCHILD
  • Created: 17/03/2021
  • Licence: CeCILL (https://cecill.info/licences.fr.html)

compute_clusters_frequency(clustering_result)

Get the frequency of each cluster present in a clustering result.

Parameters:

Name Type Description Default
clustering_result Dict[str, int]

The dictionary that contains the predicted cluster for each data ID.

 required

Returns:

Type Description
Dict[int, float]

Dict[int,float] : Frequency fo each predicted intent.
Source code in src\cognitivefactory\interactive_clustering\utils\frequency.py 
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
def compute_clusters_frequency(clustering_result: Dict[str, int]) -> Dict[int, float]:
    """
    Get the frequency of each cluster present in a clustering result.

    Args:
        clustering_result (Dict[str,int]): The dictionary that contains the predicted cluster for each data ID.

    Returns:
        Dict[int,float] : Frequency fo each predicted intent.
    """

    # Get the total number of data IDs.
    nb_of_data_IDs = len(clustering_result.keys())

    # Default case : No data, so no cluster.
    if nb_of_data_IDs == 0:
        return {}

    # Get possible clusters IDs.
    list_of_possible_cluster_IDs: List[int] = sorted(
        {clustering_result[data_ID] for data_ID in clustering_result.keys()}
    )

    # Compute frequency of clusters in `clustering_result`.
    frequence_of_clusters: Dict[int, float] = {
        cluster_ID: len([data_ID for data_ID in clustering_result if clustering_result[data_ID] == cluster_ID])
        / nb_of_data_IDs
        for cluster_ID in list_of_possible_cluster_IDs
    }

    # Return the frequence of clusters.
    return frequence_of_clusters