src.preprocessing.linear_programming

Perform the linear programming on the clusters.

Classes

LinearProgramming([min_cover, K, threshold])

Create a narrative graph from the clusters and the memberships.
class src.preprocessing.linear_programming.LinearProgramming(min_cover=0.6, K=10, threshold=0.01)[source]

Create a narrative graph from the clusters and the memberships.

Parameters:

  • min_cover (float) – the minimum coverage of the memberships.

  • K (int) – the expected length of the main story.

  • threshold (float) – remove low edges and nodes.

min_cover: float = 0.6
K: int = 10
threshold: float = 0.01
custom_transform(data, **transform_args)[source]

Create the adjacency list and weights based on the solution.

Parameters:

  • data (DataFrame) – the pandas dataframe containing the embeddings.

  • transform_args (Never) – [UNUSED] Additional keyword arguments.

Return data:

the pandas dataframe containing adjacency list and weights.

Return type:

DataFrame