renaming dataset labels as numeric values w/o gaps

renaming binary -> ucibinary
2023-10-18 11:42:52 +02:00 · 2023-10-18 10:55:56 +02:00
2 changed files with 5 additions and 0 deletions
--- a/distribution_matching/ucibinary_experiments.py
+++ b/distribution_matching/ucibinary_experiments.py
--- a/quapy/data/datasets.py
+++ b/quapy/data/datasets.py
@ -1,3 +1,6 @@
+import numpy as np
+
+
 def warn(*args, **kwargs):
    pass
 import warnings
@ -644,6 +647,8 @@ def fetch_UCIMulticlassLabelledCollection(dataset_name, data_home=None, verbose=
    else:
        data = fetch_ucirepo(id=identifier)
        X, y = data['data']['features'].to_numpy(), data['data']['targets'].to_numpy().squeeze()
+        classes = np.sort(np.unique(y))
+        y = np.searchsorted(classes, y)
        data = LabelledCollection(X, y)
        os.makedirs(os.path.dirname(file), exist_ok=True)
        with open(file, 'wb') as file:
Author	SHA1	Message	Date
Alejandro Moreo Fernandez	f08885dca3	renaming dataset labels as numeric values w/o gaps	2023-10-18 11:42:52 +02:00
Alejandro Moreo Fernandez	cfd5b00b97	renaming binary -> ucibinary	2023-10-18 10:55:56 +02:00