ResultAggregator module

Module takes result of evaluation and computes aggregated statistics.

class ResultAggregator.ResultAggregator(evaluator, statistics, measures)

Bases: object

Class takes result of evaluation and computes aggregated statistics.

aggregate()

Function triggers computation of aggregation functions for evaluation result.

compute_top_k_aggregated_measures(measures, evaluated_top_k_values)

Function computes average of given measures

compute_top_k_aggregated_measures_per_subject(measures, evaluated_top_k_values)

Function computes average of all given measures grouped by given top-k recommendation lists for each single subject and stores it in the subjects evaluation results section.

Parameters:

• measures (list) – evaluation measures to be aggregated.
• evaluated_top_k_values (list) – top-k values to be evaluated.

compute_top_k_reconstruction(evaluated_top_k_values)

Function computes reconstruction values for top-k recommendations for each subject and stores result in evaluation result dict.

Parameters:evaluated_top_k_values (list) – k-values to be evaluated for measure@k evaluation.