MAINT Parameters validation for sklearn.manifold.trustworthiness (sci…

…kit-learn#26276)

Co-authored-by: jeremie du boisberranger <[email protected]>

sklearn/manifold/_t_sne.py

@@ -27,8 +27,8 @@
 from ..neighbors import NearestNeighbors
 from ..utils import check_random_state
 from ..utils._openmp_helpers import _openmp_effective_n_threads
-from ..utils._param_validation import Interval, StrOptions
-from ..utils.validation import check_non_negative
+from ..utils._param_validation import Interval, StrOptions, validate_params
+from ..utils.validation import _num_samples, check_non_negative
 
 # mypy error: Module 'sklearn.manifold' has no attribute '_utils'
 # mypy error: Module 'sklearn.manifold' has no attribute '_barnes_hut_tsne'
@@ -446,6 +446,15 @@ def _gradient_descent(
  return p, error, i
 
 
+@validate_params(
+ {
+ "X": ["array-like", "sparse matrix"],
+ "X_embedded": ["array-like", "sparse matrix"],
+ "n_neighbors": [Interval(Integral, 1, None, closed="left")],
+ "metric": [StrOptions(set(_VALID_METRICS) | {"precomputed"}), callable],
+ },
+ prefer_skip_nested_validation=True,
+)
 def trustworthiness(X, X_embedded, *, n_neighbors=5, metric="euclidean"):
  r"""Indicate to what extent the local structure is retained.
 
@@ -504,7 +513,7 @@ def trustworthiness(X, X_embedded, *, n_neighbors=5, metric="euclidean"):
  Local Structure. Proceedings of the Twelfth International Conference on
  Artificial Intelligence and Statistics, PMLR 5:384-391, 2009.
  """
- n_samples = X.shape[0]
+ n_samples = _num_samples(X)
  if n_neighbors >= n_samples / 2:
  raise ValueError(
  f"n_neighbors ({n_neighbors}) should be less than n_samples / 2"

sklearn/tests/test_public_functions.py

@@ -202,6 +202,7 @@ def _check_function_param_validation(
  "sklearn.linear_model.orthogonal_mp",
  "sklearn.linear_model.orthogonal_mp_gram",
  "sklearn.linear_model.ridge_regression",
+ "sklearn.manifold.trustworthiness",
  "sklearn.metrics.accuracy_score",
  "sklearn.manifold.smacof",
  "sklearn.metrics.auc",