FIX Avoid fitting a pipeline without steps (scikit-learn#31723)

Author: DeaMariaLeon

sklearn/pipeline.py

@@ -320,6 +320,8 @@ def set_params(self, **kwargs):
         return self
 
     def _validate_steps(self):
+        if not self.steps:
+            raise ValueError("The pipeline is empty. Please add steps.")
         names, estimators = zip(*self.steps)
 
         # validate names
@@ -1289,7 +1291,6 @@ def __sklearn_is_fitted__(self):
 
         An empty pipeline is considered fitted.
         """
-
         # First find the last step that is not 'passthrough'
         last_step = None
         for _, estimator in reversed(self.steps):

sklearn/tests/test_pipeline.py

@@ -282,6 +282,16 @@ def test_pipeline_invalid_parameters():
     assert params == params2
 
 
+def test_empty_pipeline():
+    X = iris.data
+    y = iris.target
+
+    pipe = Pipeline([])
+    msg = "The pipeline is empty. Please add steps."
+    with pytest.raises(ValueError, match=msg):
+        pipe.fit(X, y)
+
+
 def test_pipeline_init_tuple():
     # Pipeline accepts steps as tuple
     X = np.array([[1, 2]])