classification_report_imbalanced raises opaque IndexError on target_names length mismatch

[JayeshSuryavanshi]

Describe the bug

imblearn.metrics.classification_report_imbalanced raises an opaque IndexError: list index out of range when target_names has a different length than the number of classes, whereas the scikit-learn function it mirrors — sklearn.metrics.classification_report — raises a clear, actionable ValueError on the same input.

Steps/Code to Reproduce

from imblearn.metrics import classification_report_imbalanced
from sklearn.metrics import classification_report

y_true = [0, 1, 2, 2, 2]
y_pred = [0, 0, 2, 2, 1]

classification_report_imbalanced(y_true, y_pred, target_names=["a", "b"])
# -> IndexError: list index out of range

classification_report(y_true, y_pred, target_names=["a", "b"])
# -> ValueError: Number of classes, 3, does not match size of target_names, 2.
#    Try specifying the labels parameter

Expected behavior

A clear ValueError matching scikit-learn, e.g. "Number of classes, 3, does not match size of target_names, 2. Try specifying the labels parameter".

Versions

imbalanced-learn 0.15.dev0
scikit-learn 1.9.0
Python 3.12

I'll open a PR with the fix.

[bishtashish708]

Opened #1193 with a fix.

The change mirrors sklearn's validation pattern exactly: track labels_given = labels is not None before resolving the labels array, then check for a length mismatch between labels and target_names:

• if labels was inferred from data → raise ValueError with the clear message from the issue
• if labels was explicitly provided by the caller → warnings.warn and continue (matching sklearn's behaviour for that case)

Two tests added covering both branches, all 53 existing tests still pass.