API

class BDISchemaMatchingHeatMap(*args: Any, **kwargs: Any)

BDISchemaMatchingHeatMap is a class for generating and visualizing schema matching heatmaps between a source DataFrame and a target DataFrame or predefined dataset.

Parameters:

  • source (pd.DataFrame) – The source DataFrame containing the columns to be matched.

  • target (Union[pd.DataFrame, str]) – The target DataFrame or a string identifier for a predefined dataset (default is “gdc”).

  • top_k (int) – The number of top matches to consider for each source column (default is 10).

  • heatmap_recommendations (Optional[List[Dict]]) – Optional precomputed heatmap recommendations (default is None).

  • max_chars_samples (int) – Maximum number of characters for sample values (default is 150).

  • height (int) – Height of the heatmap visualization (default is 600).

  • ai_assitant (bool) – Flag to enable AI assistant for recommendations (default is False).

plot_heatmap() → panel.Column

Plot the heatmap for the user to interact with.

get_recommendations(source: pandas.DataFrame, target: pandas.DataFrame, top_k: int) → List[bdikit.mapping_algorithms.column_mapping.topk_matchers.TopkMatching]

Get the edited recommendations based on the user interactions.

Parameters:

  • source (pd.DataFrame) – The source DataFrame.

  • target (pd.DataFrame) – The target DataFrame.

  • top_k (int) – The number of top-k recommendations to return.

Returns:

The top-k recommendations.

Return type:

List[TopkMatching]