Spaces:
Sleeping
Sleeping
File size: 3,358 Bytes
671ff9d 509822c 671ff9d 509822c 671ff9d 509822c 671ff9d 509822c bdeb145 7c8640b 671ff9d 509822c 671ff9d 3002aa2 bdeb145 671ff9d bdeb145 671ff9d 509822c 671ff9d 509822c 671ff9d 509822c 671ff9d 509822c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
import datasets
import evaluate
from sklearn.metrics import (
adjusted_mutual_info_score,
adjusted_rand_score,
calinski_harabasz_score,
completeness_score,
davies_bouldin_score,
fowlkes_mallows_score,
homogeneity_score,
silhouette_score,
)
from sklearn.metrics.cluster import contingency_matrix, pair_confusion_matrix
_CITATION = """
@article{scikit-learn,
title={Scikit-learn: Machine Learning in {P}ython},
author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V.
and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P.
and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and
Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.},
journal={Journal of Machine Learning Research},
volume={12},
pages={2825--2830},
year={2011}
}
"""
_DESCRIPTION = """\
This evaluator computes multiple clustering metrics to assess the quality of a clustering.
By default, the evaluator works as in an unsupervised setting, evaluating the clustering just from
the samples and the predictions.
However, it allows to compute additional metrics when truth labels are passed too, which is not shown in this demo.
"""
_KWARGS_DESCRIPTION = """
Computes the quality of clustering results.
Args:
samples: vector representations
predictions: predicted cluster labels
truth_labels (optional): truth labels to compute additional metrics
Returns:
silhouete_score
davies_bouldin_score
calinski_harabasz_score
completeness_score
davies_bouldin_score
fowlkes_mallows_score
homogeneity_score
silhouette_score
contingency_matrix
pair_confusion_matrix
"""
@evaluate.utils.file_utils.add_start_docstrings(
_DESCRIPTION, _KWARGS_DESCRIPTION
)
class ClusteringEvaluator(evaluate.Metric):
def _info(self):
return evaluate.MetricInfo(
module_type="metric",
description=_DESCRIPTION,
citation=_CITATION,
inputs_description=_KWARGS_DESCRIPTION,
features=datasets.Features(
{
"samples": datasets.Sequence(datasets.Value("float32")),
"predictions": datasets.Value("int64"),
}
),
)
def _compute(self, samples, predictions, truth_labels=None):
unsupervised_metrics = [
silhouette_score,
davies_bouldin_score,
calinski_harabasz_score,
]
supervised_metrics = [
adjusted_rand_score,
adjusted_mutual_info_score,
homogeneity_score,
completeness_score,
fowlkes_mallows_score,
contingency_matrix,
pair_confusion_matrix,
]
results = {}
# Compute unsupervised metrics always
for fn in unsupervised_metrics:
results[fn.__name__] = float(fn(samples, predictions))
# Compute supervised metrics if reference labels are passed
if truth_labels is not None:
for fn in supervised_metrics:
score = fn(truth_labels, predictions)
try:
score = float(score)
except (AttributeError, TypeError):
pass
results[fn.__name__] = score
return results
|