All Submissions Basics:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?
• Have you checked all Issues to tie the PR to a specific one?

All Submissions Cores:

• Have you added an explanation of what your changes do and why you'd like us to include them?
• Have you written new tests for your core changes, as applicable?
• Have you successfully ran tests with your changes locally?
• Does your submission pass tests, including CircleCI, Travis CI, and AppVeyor?
• Does your submission have appropriate code coverage? The cutoff threshold is 95% by Coversall.

Closes #540

What does this PR do?

Adds X = check_array(X) to decision_function() in four sklearn-wrapper models (IForest, OCSVM, LOF, GMM) that were missing input validation. This prevents the UserWarning: X has feature names, but <Model> was fitted without feature names warning when using pandas DataFrames.

Root cause: fit() already calls check_array(X), which strips DataFrame column names before fitting the underlying sklearn estimator. But decision_function() was passing the raw DataFrame directly to sklearn at predict time, causing a feature-name mismatch warning.

Three other wrapper models (MCD, PCA, HDBSCAN) already had check_array(X) in their decision_function() and were unaffected.

Changes

• pyod/models/iforest.py — add X = check_array(X) in decision_function()
• pyod/models/ocsvm.py — same
• pyod/models/lof.py — same
• pyod/models/gmm.py — same
• pyod/test/test_iforest.py — add regression test test_dataframe_no_feature_name_warning