Clustering module#

class libadalina_analytics.clustering.ClusteringDistance(*, name: str, weight: float | None = 1.0, function: str | None = None)[source]#
function: str | None#
Distance function to use. Must be one of the following:

  • euclidean

  • manhattan

  • chebyshev

  • haversine

  • hamming

  • canberra

  • braycurtis

  • jaccard

  • matching

  • dice

  • kulsinski

  • rogerstanimoto

  • russellrao

  • sokalmichener

  • sokalsneath

If None, the default function depends on the type of the values in the column:

  • geometric values default to scipy distance_matrix

  • float and int values default to euclidean

  • string values default to hamming

  • boolean values default to jaccard

model_config: ClassVar[ConfigDict] = {}#

Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

name: str#

Name of the column in the input data to be used for distance calculation.

weight: float | None#

Weight of the distance in the overall distance calculation. Default is 1.0.

libadalina_analytics.clustering.clustering_algorithm(data: DataFrame | GeoDataFrame | DataFrame, epsg: EPSGFormats, geometry_column: str = 'geometry', geometry_format: GeometryFormats = GeometryFormats.WKT, weight_column: str | None = None, k_min: int | None = None, k_max: int | None = None, f_min: float | None = None, distances: list[ClusteringDistance] | None = None, timelimit: int = 60) → AdalinaZoningSolution | None[source]#

Create clusters of similar areas minimizing their internal distance.

Distances are given as a list of ClusteringDistance objects, each containing: - name: the name of the column in the input data - weight: the weight of the distance in the overall distance calculation - function: the distance function to use

Parameters:

  • data (pandas.DataFrame or geopandas.GeoDataFrame or pyspark.sql.DataFrame) – The input data containing the geometries and attributes to be clustered.

  • epsg (EPSGFormats) – The EPSG format of the input geometries.

  • geometry_column (str) – The name of the column containing the geometries. Default is ‘geometry’.

  • geometry_format (GeometryFormats) – The format of the geometries in the geometry column. Default is GeometryFormats.WKT.

  • weight_column (str | None) – The name of the column to use as the optional weight of an area. Default is None.

  • k_min (int | None) – The minimum number of clusters to create. If None, defaults to 1.

  • k_max (int | None) – The maximum number of clusters to create. Default is None.

  • f_min (float | None) – The minimum total weight of a cluster. If None, defaults to 1.

  • distances (list[ClusteringDistance] | None) – List of distances to use for clustering. If None, no only distance between area centroids will be used. Each distance is represented as a ClusteringDistance object that includes the name of the column, the weight of the distance, and the distance function to use.

  • timelimit (int) – The maximum time (in seconds) to run the algorithm. Default is 60 seconds.

Return type:

AdalinaZoningSolution