"""
This is the metrics' module.
This module contains and expose the recommendation metrics.
Each metric is encapsulated in a specific package.
See the implementation of Precision metric for creating new per-user metrics.
See the implementation of Item Coverage for creating new cross-user metrics.
"""
__version__ = '0.3.1'
__author__ = 'Vito Walter Anelli, Claudio Pomo'
__email__ = 'vitowalter.anelli@poliba.it, claudio.pomo@poliba.it'
from elliot.evaluation.metrics.accuracy.ndcg import nDCG, nDCGRendle2020
from elliot.evaluation.metrics.accuracy.precision import Precision
from elliot.evaluation.metrics.accuracy.recall import Recall
from elliot.evaluation.metrics.accuracy.hit_rate import HR
from elliot.evaluation.metrics.accuracy.mrr import MRR
from elliot.evaluation.metrics.accuracy.map import MAP
from elliot.evaluation.metrics.accuracy.mar import MAR
from elliot.evaluation.metrics.accuracy.f1 import F1, ExtendedF1
from elliot.evaluation.metrics.accuracy.DSC import DSC
from elliot.evaluation.metrics.accuracy.AUC import LAUC, AUC, GAUC
from elliot.evaluation.metrics.rating.mae import MAE
from elliot.evaluation.metrics.rating.mse import MSE
from elliot.evaluation.metrics.rating.rmse import RMSE
from elliot.evaluation.metrics.coverage import ItemCoverage, UserCoverage, NumRetrieved, UserCoverageAtN
from elliot.evaluation.metrics.diversity.gini_index import GiniIndex
from elliot.evaluation.metrics.diversity.shannon_entropy import ShannonEntropy
from elliot.evaluation.metrics.diversity.SRecall import SRecall
from elliot.evaluation.metrics.novelty.EFD import EFD, ExtendedEFD
from elliot.evaluation.metrics.novelty.EPC import EPC, ExtendedEPC
from elliot.evaluation.metrics.bias import ARP, APLT, ACLT, PopRSP, PopREO, ExtendedPopRSP, ExtendedPopREO
from elliot.evaluation.metrics.fairness.MAD import UserMADrating, ItemMADrating, UserMADranking, ItemMADranking
from elliot.evaluation.metrics.fairness.BiasDisparity import BiasDisparityBR, BiasDisparityBS, BiasDisparityBD
from elliot.evaluation.metrics.fairness.rsp import RSP
from elliot.evaluation.metrics.fairness.reo import REO
from elliot.evaluation.metrics.statistical_array_metric import StatisticalMetric
_metric_dictionary = {
"nDCG": nDCG,
"nDCGRendle2020": nDCGRendle2020,
"Precision": Precision,
"Recall": Recall,
"HR": HR,
"MRR": MRR,
"MAP": MAP,
"MAR": MAR,
"F1": F1,
"ExtendedF1": ExtendedF1,
"DSC": DSC,
"LAUC": LAUC,
"GAUC": GAUC,
"AUC": AUC,
"ItemCoverage": ItemCoverage,
"UserCoverage": UserCoverage,
"UserCoverageAtN": UserCoverageAtN,
"NumRetrieved": NumRetrieved,
"Gini": GiniIndex,
"SEntropy": ShannonEntropy,
"EFD": EFD,
"ExtendedEFD": ExtendedEFD,
"EPC": EPC,
"ExtendedEPC": ExtendedEPC,
"MAE": MAE,
"MSE": MSE,
"RMSE": RMSE,
"UserMADrating": UserMADrating,
"ItemMADrating": ItemMADrating,
"UserMADranking": UserMADranking,
"ItemMADranking": ItemMADranking,
"BiasDisparityBR": BiasDisparityBR,
"BiasDisparityBS": BiasDisparityBS,
"BiasDisparityBD": BiasDisparityBD,
"SRecall": SRecall,
"ARP": ARP,
"APLT": APLT,
"ACLT": ACLT,
"PopRSP": PopRSP,
"PopREO": PopREO,
"ExtendedPopRSP": ExtendedPopRSP,
"ExtendedPopREO": ExtendedPopREO,
"RSP": RSP,
"REO": REO
}
_lower_dict = {k.lower(): v for k, v in _metric_dictionary.items()}
[docs]def parse_metrics(metrics):
return [_lower_dict[m.lower()] for m in metrics if m.lower() in _lower_dict.keys()]
[docs]def parse_metric(metric):
metric = metric.lower()
return _lower_dict[metric] if metric in _lower_dict.keys() else None