MAINT Parameters validation for sklearn.random_projection.johnson_lindenstrauss_min_dim (scikit-learn#25278)

DeaMariaLeon · web-flow · commit 38d7fba5866f · 2023-01-06T13:11:50.000+01:00
diff --git a/sklearn/random_projection.py b/sklearn/random_projection.py
@@ -38,7 +38,7 @@
 from .base import ClassNamePrefixFeaturesOutMixin
 
 from .utils import check_random_state
-from .utils._param_validation import Interval, StrOptions
+from .utils._param_validation import Interval, StrOptions, validate_params
 from .utils.extmath import safe_sparse_dot
 from .utils.random import sample_without_replacement
 from .utils.validation import check_array, check_is_fitted
@@ -51,11 +51,17 @@
 ]
 
 
+@validate_params(
+    {
+        "n_samples": ["array-like", Interval(Real, 1, None, closed="left")],
+        "eps": ["array-like", Interval(Real, 0, 1, closed="neither")],
+    }
+)
 def johnson_lindenstrauss_min_dim(n_samples, *, eps=0.1):
     """Find a 'safe' number of components to randomly project to.
 
     The distortion introduced by a random projection `p` only changes the
-    distance between two points by a factor (1 +- eps) in an euclidean space
+    distance between two points by a factor (1 +- eps) in a euclidean space
     with good probability. The projection `p` is an eps-embedding as defined
     by:
 
@@ -81,12 +87,12 @@ def johnson_lindenstrauss_min_dim(n_samples, *, eps=0.1):
     Parameters
     ----------
     n_samples : int or array-like of int
-        Number of samples that should be a integer greater than 0. If an array
+        Number of samples that should be an integer greater than 0. If an array
         is given, it will compute a safe number of components array-wise.
 
-    eps : float or ndarray of shape (n_components,), dtype=float, \
+    eps : float or array-like of shape (n_components,), dtype=float, \
             default=0.1
-        Maximum distortion rate in the range (0,1 ) as defined by the
+        Maximum distortion rate in the range (0, 1) as defined by the
         Johnson-Lindenstrauss lemma. If an array is given, it will compute a
         safe number of components array-wise.
 
@@ -123,7 +129,7 @@ def johnson_lindenstrauss_min_dim(n_samples, *, eps=0.1):
     if np.any(eps <= 0.0) or np.any(eps >= 1):
         raise ValueError("The JL bound is defined for eps in ]0, 1[, got %r" % eps)
 
-    if np.any(n_samples) <= 0:
+    if np.any(n_samples <= 0):
         raise ValueError(
             "The JL bound is defined for n_samples greater than zero, got %r"
             % n_samples
diff --git a/sklearn/tests/test_public_functions.py b/sklearn/tests/test_public_functions.py
@@ -120,6 +120,7 @@ def _check_function_param_validation(
     "sklearn.metrics.roc_curve",
     "sklearn.metrics.zero_one_loss",
     "sklearn.model_selection.train_test_split",
+    "sklearn.random_projection.johnson_lindenstrauss_min_dim",
     "sklearn.svm.l1_min_c",
 ]
 
diff --git a/sklearn/tests/test_random_projection.py b/sklearn/tests/test_random_projection.py
@@ -65,7 +65,12 @@ def densify(matrix):
 
 
 @pytest.mark.parametrize(
-    "n_samples, eps", [(100, 1.1), (100, 0.0), (100, -0.1), (0, 0.5)]
+    "n_samples, eps",
+    [
+        ([100, 110], [0.9, 1.1]),
+        ([90, 100], [0.1, 0.0]),
+        ([50, -40], [0.1, 0.2]),
+    ],
 )
 def test_invalid_jl_domain(n_samples, eps):
     with pytest.raises(ValueError):

Original file line number	Diff line number	Diff line change
`@@ -120,6 +120,7 @@ def _check_function_param_validation(`
`120`	`120`	`"sklearn.metrics.roc_curve",`
`121`	`121`	`"sklearn.metrics.zero_one_loss",`
`122`	`122`	`"sklearn.model_selection.train_test_split",`
	`123`	`+ "sklearn.random_projection.johnson_lindenstrauss_min_dim",`
`123`	`124`	`"sklearn.svm.l1_min_c",`
`124`	`125`	`]`
`125`	`126`