Feature range parameter support - Unit Tests

codebeaver.yml

@@ -0,0 +1,2 @@
+from: pytest
+# This file was generated automatically by CodeBeaver based on your repository. Learn how to customize it here: https://docs.codebeaver.ai/configuration/

mlxtend/_base/tests/test_base_model.py

@@ -1,4 +1,4 @@
-# Sebastian Raschka 2014-2024
+# Sebastian Raschka 2014-2023
 # mlxtend Machine Learning Library Extensions
 # Author: Sebastian Raschka <sebastianraschka.com>
 #
@@ -8,10 +8,13 @@
 
 from mlxtend._base import _BaseModel
 from mlxtend.utils import assert_raises
+import pytest
+import time
 
 
 class BlankModel(_BaseModel):
     def __init__(self, print_progress=0, random_seed=1):
+        super().__init__()
         self.print_progress = print_progress
         self.random_seed = random_seed
 
@@ -44,3 +47,184 @@ def test_check_array_3():
     X = np.array([[1], [2], [3]])
     est = BlankModel(print_progress=0, random_seed=1)
     est._check_arrays(X)
+
+
+
+def test_get_params():
+    """Test that get_params returns the expected parameter dictionary for BlankModel."""
+    est = BlankModel(print_progress=5, random_seed=42)
+    params = est.get_params()
+    assert 'print_progress' in params
+    assert 'random_seed' in params
+    assert params['print_progress'] == 5
+    assert params['random_seed'] == 42
+
+
+def test_set_params():
+    """Test that set_params updates the parameters in the estimator."""
+    est = BlankModel(print_progress=0, random_seed=1)
+    est.set_params(print_progress=10, random_seed=99)
+    params = est.get_params()
+    assert params['print_progress'] == 10
+    assert params['random_seed'] == 99
+
+
+# Dummy submodel for testing nested parameter access and update
+class DummySubModel(object):
+    def __init__(self, alpha=0.1):
+        self.alpha = alpha
+    def get_params(self, deep=True):
+        return {'alpha': self.alpha}
+    def set_params(self, **params):
+        if 'alpha' in params:
+            self.alpha = params['alpha']
+        return self
+
+
+# Dummy estimator to test nested parameters functionality in _BaseModel
+class DummyEstimator(_BaseModel):
+    def __init__(self, sub_model=None, gamma=1.0):
+        self.sub_model = sub_model if sub_model is not None else DummySubModel()
+        self.gamma = gamma
+
+
+def test_get_and_set_nested_params():
+    """Test that get_params nests parameters correctly and that set_params updates nested parameters."""
+    est = DummyEstimator(gamma=1.0)
+    params = est.get_params()
+    # Expect nested key 'sub_model__alpha' along with 'gamma'
+    assert 'gamma' in params
+    assert 'sub_model__alpha' in params
+    assert params['gamma'] == 1.0
+    assert params['sub_model__alpha'] == 0.1
+
+    # Now update nested parameter and gamma via set_params
+    est.set_params(sub_model__alpha=0.5, gamma=2.0)
+    params = est.get_params()
+    assert params['gamma'] == 2.0
+    assert params['sub_model__alpha'] == 0.5
+
+
+def test_check_arrays_mismatched_y():
+    """Test that _check_arrays raises ValueError when the number of samples in X and y do not match."""
+    X = np.array([[1], [2], [3]])
+    y = np.array([1, 2])  # mismatched length: 3 vs 2
+    est = BlankModel(print_progress=0, random_seed=1)
+    assert_raises(ValueError, "X and y must contain the same number of samples", est._check_arrays, X, y)
+
+
+def test_bad_get_param_names():
+    """Test that _get_param_names raises an error for models with variable positional arguments."""
+    class BadModel(_BaseModel):
+        def __init__(self, *args, param=0):
+            pass
+    with pytest.raises(RuntimeError, match="scikit-learn estimators should always specify their parameters"):
+        _ = BadModel._get_param_names()
+def test_init_time():
+    """Test that _init_time attribute is set upon initialization with a valid timestamp."""
+    est = BlankModel()
+    # Verify that _init_time exists and is a recent timestamp (within the last 5 seconds)
+    assert hasattr(est, "_init_time")
+    assert time.time() - est._init_time < 5
+
+def test_set_params_empty():
+    """Test that calling set_params with no parameters does not change the estimator."""
+    est = BlankModel(print_progress=3, random_seed=7)
+    params_before = est.get_params()
+    est.set_params()
+    params_after = est.get_params()
+    assert params_before == params_after
+
+def test_check_arrays_y_invalid():
+    """Test that passing a non-array y (without a shape attribute) raises a TypeError."""
+    X = np.array([[1], [2], [3]])
+    y = 5  # invalid y, not a numpy array
+    est = BlankModel()
+    with pytest.raises(TypeError):
+        est._check_arrays(X, y)
+
+def test_get_param_names_on_base():
+    """Test that _get_param_names returns an empty list for _BaseModel since its __init__ takes only self."""
+    param_names = _BaseModel._get_param_names()
+    assert param_names == []
+
+def test_invalid_param():
+    """Test that setting an invalid parameter via set_params raises a ValueError."""
+    est = BlankModel(print_progress=0, random_seed=1)
+    with pytest.raises(ValueError, match="Invalid parameter"):
+        est.set_params(nonexistent_param=123)
+
+def test_get_params_deep_false():
+    """Test that get_params with deep=False does not expand nested estimator parameters."""
+    # Create a DummyEstimator to test nested parameter expansion
+    est = DummyEstimator(gamma=3.0)
+    params = est.get_params(deep=False)
+    # The deep expansion should not be performed; thus no 'sub_model__alpha' key should be present,
+    # although the 'sub_model' key (the nested object) should be.
+    assert "sub_model__alpha" not in params
+    assert "sub_model" in params
+    assert params["gamma"] == 3.0
+def test_get_param_names_custom():
+    """Test that _get_param_names returns parameter names for a custom model."""
+    class CustomModel(_BaseModel):
+        def __init__(self, param1=1, param2=2):
+            super().__init__()
+            self.param1 = param1
+            self.param2 = param2
+
+    expected = ['param1', 'param2']
+    param_names = CustomModel._get_param_names()
+    assert param_names == sorted(expected)
+
+def test_set_params_non_estimator_nested():
+    """Test that set_params works when a nested parameter does not have get_params."""
+    class DummyNonEstimator:
+        def __init__(self, val=100):
+            self.val = val
+
+    class ModelWithNonEstimator(_BaseModel):
+        def __init__(self, dummy=None):
+            super().__init__()
+            self.dummy = dummy if dummy is not None else DummyNonEstimator()
+
+    model = ModelWithNonEstimator()
+    params = model.get_params()
+    # There should be no nested keys since dummy doesn't implement get_params
+    assert "dummy__val" not in params
+    assert "dummy" in params
+
+    # Now update dummy by passing a new DummyNonEstimator instance
+    new_dummy = DummyNonEstimator(val=200)
+    model.set_params(dummy=new_dummy)
+    new_params = model.get_params()
+    assert new_params["dummy"].val == 200
+
+def test_set_nested_invalid_param_no_error():
+    """Test that setting a nested parameter that does not exist in the submodel is silently ignored."""
+    sub = DummySubModel(alpha=0.1)
+    est = DummyEstimator(sub_model=sub, gamma=1.0)
+    # Attempt to set a nested parameter that doesn't exist; the submodel should ignore it
+    est.set_params(sub_model__nonexistent=123)
+    params = est.get_params()
+    # The valid parameter should remain unchanged
+    assert params['sub_model__alpha'] == 0.1
+def test_get_param_names_with_deprecated():
+    """Test that _get_param_names retrieves parameters from a deprecated_original __init__ if available."""
+    class DeprecatedModel(_BaseModel):
+        def __init__(self, a=1, b=2):
+            super().__init__()
+            self.a = a
+            self.b = b
+
+    # Define a dummy deprecated __init__ with the intended signature.
+    def dummy_init(self, a=1, b=2):
+        super(DeprecatedModel, self).__init__()
+        self.a = a
+        self.b = b
+
+    # Attach dummy_init as the deprecated_original.
+    DeprecatedModel.__init__.deprecated_original = dummy_init
+
+    param_names = DeprecatedModel._get_param_names()
+    # Expected sorted list of parameters ['a', 'b']
+    assert param_names == sorted(['a', 'b'])

mlxtend/classifier/tests/test_adaline.py

@@ -1,4 +1,4 @@
-# Sebastian Raschka 2014-2024
+# Sebastian Raschka 2014-2023
 # mlxtend Machine Learning Library Extensions
 # Author: Sebastian Raschka <sebastianraschka.com>
 #
@@ -34,7 +34,7 @@ def test_invalid_labels_1():
     else:
         objtype = "set([(0, 1)])"
 
-    expect = "Labels not in %s.\nFound (1, 2)" % objtype
+    expect = "Labels not in %s.\nFound (np.int64(1), np.int64(2))" % objtype
 
     assert_raises(AttributeError, expect, ada.fit, X, y2, {(0, 1)})
 

mlxtend/classifier/tests/test_logistic_regression.py

@@ -1,4 +1,4 @@
-# Sebastian Raschka 2014-2024
+# Sebastian Raschka 2014-2023
 # mlxtend Machine Learning Library Extensions
 # Author: Sebastian Raschka <sebastianraschka.com>
 #
@@ -32,7 +32,7 @@ def test_invalid_labels_1():
     else:
         objtype = "set([(0, 1)])"
 
-    expect = "Labels not in %s.\nFound (1, 2)" % objtype
+    expect = "Labels not in %s.\nFound (np.int64(1), np.int64(2))" % objtype
 
     assert_raises(AttributeError, expect, lr.fit, X, y1, {(0, 1)})
 

mlxtend/classifier/tests/test_perceptron.py

@@ -1,4 +1,4 @@
-# Sebastian Raschka 2014-2024
+# Sebastian Raschka 2014-2023
 # mlxtend Machine Learning Library Extensions
 # Author: Sebastian Raschka <sebastianraschka.com>
 #
@@ -34,7 +34,7 @@ def test_invalid_labels_1():
     else:
         objtype = "set([(0, 1)])"
 
-    expect = "Labels not in %s.\nFound (1, 2)" % objtype
+    expect = "Labels not in %s.\nFound (np.int64(1), np.int64(2))" % objtype
 
     assert_raises(AttributeError, expect, ppn.fit, X, y1, {(0, 1)})
 

mlxtend/evaluate/tests/test_bias_variance_decomp.py

@@ -1,4 +1,4 @@
-# Sebastian Raschka 2014-2024
+# Sebastian Raschka 2014-2023
 # mlxtend Machine Learning Library Extensions
 #
 # Nonparametric Permutation Test
@@ -113,14 +113,9 @@ def test_mse_bagging():
     APPVEYOR = False
 
 
-GITHUB_ACTIONS = os.getenv("GITHUB_ACTIONS_CI", "false").lower() == "true"
-
-
-@pytest.mark.skipif(
-    TRAVIS or APPVEYOR or GITHUB_ACTIONS, reason="TensorFlow dependency"
-)
+@pytest.mark.skipif(TRAVIS or APPVEYOR, reason="TensorFlow dependency")
 def test_keras():
-    import tensorflow as tf
+    tf = pytest.importorskip("tensorflow")
 
     X, y = boston_housing_data()
     X_train, X_test, y_train, y_test = train_test_split(