From 2ca67a6e5de506f2c155ee2bf971405398615684 Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Thu, 26 May 2022 10:50:35 -0400
Subject: [PATCH 01/15] Reconciled changelog

---
 doc/whats_new/v1.1.rst | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/doc/whats_new/v1.1.rst b/doc/whats_new/v1.1.rst
index 5a5e3bc71bcc7..4c46c0d631f76 100644
--- a/doc/whats_new/v1.1.rst
+++ b/doc/whats_new/v1.1.rst
@@ -2,6 +2,23 @@
 
 .. currentmodule:: sklearn
 
+.. _changes_1_1_2:
+
+Version 1.1.2
+=============
+
+**In Development**
+
+Changelog
+---------
+
+:mod:`sklearn.cluster`
+......................
+
+- |Fix| Fixed a bug in :class:`cluster.Birch` that could trigger an error when splitting
+  a node if there are duplicates in the dataset.
+  :pr:`23395` by :user:`Jérémie du Boisberranger <jeremiedbb>`.
+
 .. _changes_1_1_1:
 
 Version 1.1.1

From 15024ac57e9ef0e5b2e9b4576184fd111060a391 Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Thu, 26 May 2022 11:13:11 -0400
Subject: [PATCH 02/15] Completed initial deprecation

---
 sklearn/cluster/_affinity_propagation.py      | 51 ++++++++++++-------
 .../tests/test_affinity_propagation.py        | 47 ++++++++++++-----
 2 files changed, 69 insertions(+), 29 deletions(-)

diff --git a/sklearn/cluster/_affinity_propagation.py b/sklearn/cluster/_affinity_propagation.py
index f0274b113a341..3282724291761 100644
--- a/sklearn/cluster/_affinity_propagation.py
+++ b/sklearn/cluster/_affinity_propagation.py
@@ -302,10 +302,14 @@ class AffinityPropagation(ClusterMixin, BaseEstimator):
         they will be set to the median of the input similarities.
 
     affinity : {'euclidean', 'precomputed'}, default='euclidean'
-        Which affinity to use. At the moment 'precomputed' and
-        ``euclidean`` are supported. 'euclidean' uses the
+        Which metric to use. At the moment 'precomputed' and
+        'euclidean' are supported. 'euclidean' uses the
         negative squared euclidean distance between points.
 
+    .. deprecated:: 1.2
+       `affinity` was deprecated in version 1.2 and will be renamed to `metric`
+        in 1.4.
+
     verbose : bool, default=False
         Whether to be verbose.
 
@@ -413,7 +417,8 @@ def __init__(
         convergence_iter=15,
         copy=True,
         preference=None,
-        affinity="euclidean",
+        affinity="deprecated",  # TODO(1.4): Remove
+        metric=None,  # TODO(1.4): Set default as "euclidean"
         verbose=False,
         random_state=None,
     ):
@@ -425,21 +430,22 @@ def __init__(
         self.verbose = verbose
         self.preference = preference
         self.affinity = affinity
+        self.metric = metric
         self.random_state = random_state
 
     def _more_tags(self):
-        return {"pairwise": self.affinity == "precomputed"}
+        return {"pairwise": self.metric == "precomputed"}
 
     def fit(self, X, y=None):
-        """Fit the clustering from features, or affinity matrix.
+        """Fit the clustering from features, or distance matrix.
 
         Parameters
         ----------
         X : {array-like, sparse matrix} of shape (n_samples, n_features), or \
                 array-like of shape (n_samples, n_samples)
-            Training instances to cluster, or similarities / affinities between
-            instances if ``affinity='precomputed'``. If a sparse feature matrix
-            is provided, it will be converted into a sparse ``csr_matrix``.
+            Training instances to cluster, or distances between instances if
+            ``metric='precomputed'``. If a sparse feature matrix is provided,
+            it will be converted into a sparse ``csr_matrix``.
 
         y : Ignored
             Not used, present here for API consistency by convention.
@@ -449,19 +455,30 @@ def fit(self, X, y=None):
         self
             Returns the instance itself.
         """
-        if self.affinity == "precomputed":
+        # TODO(1.4): Remove
+        self._metric = self.metric or self.affinity
+        if self.affinity != "deprecated":
+            warnings.warn(
+                "Attribute `affinity` was deprecated in version 1.2 and will be removed"
+                " in 1.4. Use `metric` instead",
+                DeprecationWarning,
+            )
+        elif self._metric == "deprecated":
+            self._metric = "euclidean"
+
+        if self._metric == "precomputed":
             accept_sparse = False
         else:
             accept_sparse = "csr"
         X = self._validate_data(X, accept_sparse=accept_sparse)
-        if self.affinity == "precomputed":
+        if self._metric == "precomputed":
             self.affinity_matrix_ = X
-        elif self.affinity == "euclidean":
+        elif self._metric == "euclidean":
             self.affinity_matrix_ = -euclidean_distances(X, squared=True)
         else:
             raise ValueError(
-                "Affinity must be 'precomputed' or 'euclidean'. Got %s instead"
-                % str(self.affinity)
+                "metric must be 'precomputed' or 'euclidean'. Got %s instead"
+                % str(self._metric)
             )
 
         check_scalar(
@@ -496,7 +513,7 @@ def fit(self, X, y=None):
             random_state=self.random_state,
         )
 
-        if self.affinity != "precomputed":
+        if self.metric != "precomputed":
             self.cluster_centers_ = X[self.cluster_centers_indices_].copy()
 
         return self
@@ -519,7 +536,7 @@ def predict(self, X):
         X = self._validate_data(X, reset=False, accept_sparse="csr")
         if not hasattr(self, "cluster_centers_"):
             raise ValueError(
-                "Predict method is not supported when affinity='precomputed'."
+                "Predict method is not supported when metric='precomputed'."
             )
 
         if self.cluster_centers_.shape[0] > 0:
@@ -535,14 +552,14 @@ def predict(self, X):
             return np.array([-1] * X.shape[0])
 
     def fit_predict(self, X, y=None):
-        """Fit clustering from features/affinity matrix; return cluster labels.
+        """Fit clustering from features/distance matrix; return cluster labels.
 
         Parameters
         ----------
         X : {array-like, sparse matrix} of shape (n_samples, n_features), or \
                 array-like of shape (n_samples, n_samples)
             Training instances to cluster, or similarities / affinities between
-            instances if ``affinity='precomputed'``. If a sparse feature matrix
+            instances if ``metric='precomputed'``. If a sparse feature matrix
             is provided, it will be converted into a sparse ``csr_matrix``.
 
         y : Ignored
diff --git a/sklearn/cluster/tests/test_affinity_propagation.py b/sklearn/cluster/tests/test_affinity_propagation.py
index e5dc5d584266d..57012ab5b69b2 100644
--- a/sklearn/cluster/tests/test_affinity_propagation.py
+++ b/sklearn/cluster/tests/test_affinity_propagation.py
@@ -45,7 +45,7 @@ def test_affinity_propagation():
     assert n_clusters == n_clusters_
 
     af = AffinityPropagation(
-        preference=preference, affinity="precomputed", random_state=28
+        preference=preference, metric="precomputed", random_state=28
     )
     labels_precomputed = af.fit(S).labels_
 
@@ -82,10 +82,11 @@ def test_affinity_propagation_affinity_shape():
         (X, {"damping": 2}, ValueError, "damping == 2, must be < 1"),
         (X, {"max_iter": 0}, ValueError, "max_iter == 0, must be >= 1."),
         (X, {"convergence_iter": 0}, ValueError, "convergence_iter == 0, must be >= 1"),
-        (X, {"affinity": "unknown"}, ValueError, "Affinity must be"),
+        # TODO(1.4): Replace message "Affinity must be" with "metric must be"
+        (X, {"metric": "unknown"}, ValueError, "metric must be"),
         (
             csr_matrix((3, 3)),
-            {"affinity": "precomputed"},
+            {"metric": "precomputed"},
             TypeError,
             "A sparse matrix was passed, but dense data is required",
         ),
@@ -99,7 +100,7 @@ def test_affinity_propagation_params_validation(input, params, err_type, err_msg
 
 def test_affinity_propagation_predict():
     # Test AffinityPropagation.predict
-    af = AffinityPropagation(affinity="euclidean", random_state=63)
+    af = AffinityPropagation(metric="euclidean", random_state=63)
     labels = af.fit_predict(X)
     labels2 = af.predict(X)
     assert_array_equal(labels, labels2)
@@ -108,13 +109,13 @@ def test_affinity_propagation_predict():
 def test_affinity_propagation_predict_error():
     # Test exception in AffinityPropagation.predict
     # Not fitted.
-    af = AffinityPropagation(affinity="euclidean")
+    af = AffinityPropagation(metric="euclidean")
     with pytest.raises(ValueError):
         af.predict(X)
 
     # Predict not supported when affinity="precomputed".
     S = np.dot(X, X.T)
-    af = AffinityPropagation(affinity="precomputed", random_state=57)
+    af = AffinityPropagation(metric="precomputed", random_state=57)
     af.fit(S)
     with pytest.raises(ValueError):
         af.predict(X)
@@ -186,7 +187,7 @@ def test_affinity_propagation_predict_non_convergence():
 
 def test_affinity_propagation_non_convergence_regressiontest():
     X = np.array([[1, 0, 0, 0, 0, 0], [0, 1, 1, 1, 0, 0], [0, 0, 1, 0, 0, 1]])
-    af = AffinityPropagation(affinity="euclidean", max_iter=2, random_state=34)
+    af = AffinityPropagation(metric="euclidean", max_iter=2, random_state=34)
     msg = (
         "Affinity propagation did not converge, this model may return degenerate"
         " cluster centers and labels."
@@ -258,9 +259,7 @@ def test_affinity_propagation_float32():
     X = np.array(
         [[1, 0, 0, 0], [0, 1, 1, 0], [0, 1, 1, 0], [0, 0, 0, 1]], dtype="float32"
     )
-    afp = AffinityPropagation(preference=1, affinity="precomputed", random_state=0).fit(
-        X
-    )
+    afp = AffinityPropagation(preference=1, metric="precomputed", random_state=0).fit(X)
     expected = np.array([0, 1, 1, 2])
     assert_array_equal(afp.labels_, expected)
 
@@ -268,7 +267,7 @@ def test_affinity_propagation_float32():
 def test_sparse_input_for_predict():
     # Test to make sure sparse inputs are accepted for predict
     # (non-regression test for issue #20049)
-    af = AffinityPropagation(affinity="euclidean", random_state=42)
+    af = AffinityPropagation(metric="euclidean", random_state=42)
     af.fit(X)
     labels = af.predict(csr_matrix((2, 2)))
     assert_array_equal(labels, (2, 2))
@@ -277,8 +276,32 @@ def test_sparse_input_for_predict():
 def test_sparse_input_for_fit_predict():
     # Test to make sure sparse inputs are accepted for fit_predict
     # (non-regression test for issue #20049)
-    af = AffinityPropagation(affinity="euclidean", random_state=42)
+    af = AffinityPropagation(metric="euclidean", random_state=42)
     rng = np.random.RandomState(42)
     X = csr_matrix(rng.randint(0, 2, size=(5, 5)))
     labels = af.fit_predict(X)
     assert_array_equal(labels, (0, 1, 1, 2, 3))
+
+
+# TODO(1.4): Remove
+def test_deprecate_affinity():
+    est = [
+        AffinityPropagation(random_state=42),
+        AffinityPropagation(metric="euclidean", random_state=42),
+    ]
+    with warnings.catch_warnings():
+        warnings.simplefilter("error", DeprecationWarning)
+        for af in est:
+            af.fit(X)
+            af.predict(X)
+            af.fit_predict(X)
+
+    af = AffinityPropagation(affinity="euclidean", random_state=42)
+    msg = (
+        "Attribute `affinity` was deprecated in version 1.2 and will be removed in 1.4."
+        " Use `metric` instead"
+    )
+    with pytest.warns(DeprecationWarning, match=msg):
+        af.fit(X)
+        af.predict(X)
+        af.fit_predict(X)

From 59d7ce6675843a566a74d1019b48804d3b7708f9 Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Thu, 26 May 2022 11:33:34 -0400
Subject: [PATCH 03/15] Fixed deprecation

---
 sklearn/cluster/_affinity_propagation.py      | 51 +++++++------------
 sklearn/cluster/_agglomerative.py             | 35 ++++++++++---
 .../tests/test_affinity_propagation.py        | 47 +++++------------
 sklearn/cluster/tests/test_hierarchical.py    | 37 +++++++++++---
 4 files changed, 89 insertions(+), 81 deletions(-)

diff --git a/sklearn/cluster/_affinity_propagation.py b/sklearn/cluster/_affinity_propagation.py
index 3282724291761..f0274b113a341 100644
--- a/sklearn/cluster/_affinity_propagation.py
+++ b/sklearn/cluster/_affinity_propagation.py
@@ -302,14 +302,10 @@ class AffinityPropagation(ClusterMixin, BaseEstimator):
         they will be set to the median of the input similarities.
 
     affinity : {'euclidean', 'precomputed'}, default='euclidean'
-        Which metric to use. At the moment 'precomputed' and
-        'euclidean' are supported. 'euclidean' uses the
+        Which affinity to use. At the moment 'precomputed' and
+        ``euclidean`` are supported. 'euclidean' uses the
         negative squared euclidean distance between points.
 
-    .. deprecated:: 1.2
-       `affinity` was deprecated in version 1.2 and will be renamed to `metric`
-        in 1.4.
-
     verbose : bool, default=False
         Whether to be verbose.
 
@@ -417,8 +413,7 @@ def __init__(
         convergence_iter=15,
         copy=True,
         preference=None,
-        affinity="deprecated",  # TODO(1.4): Remove
-        metric=None,  # TODO(1.4): Set default as "euclidean"
+        affinity="euclidean",
         verbose=False,
         random_state=None,
     ):
@@ -430,22 +425,21 @@ def __init__(
         self.verbose = verbose
         self.preference = preference
         self.affinity = affinity
-        self.metric = metric
         self.random_state = random_state
 
     def _more_tags(self):
-        return {"pairwise": self.metric == "precomputed"}
+        return {"pairwise": self.affinity == "precomputed"}
 
     def fit(self, X, y=None):
-        """Fit the clustering from features, or distance matrix.
+        """Fit the clustering from features, or affinity matrix.
 
         Parameters
         ----------
         X : {array-like, sparse matrix} of shape (n_samples, n_features), or \
                 array-like of shape (n_samples, n_samples)
-            Training instances to cluster, or distances between instances if
-            ``metric='precomputed'``. If a sparse feature matrix is provided,
-            it will be converted into a sparse ``csr_matrix``.
+            Training instances to cluster, or similarities / affinities between
+            instances if ``affinity='precomputed'``. If a sparse feature matrix
+            is provided, it will be converted into a sparse ``csr_matrix``.
 
         y : Ignored
             Not used, present here for API consistency by convention.
@@ -455,30 +449,19 @@ def fit(self, X, y=None):
         self
             Returns the instance itself.
         """
-        # TODO(1.4): Remove
-        self._metric = self.metric or self.affinity
-        if self.affinity != "deprecated":
-            warnings.warn(
-                "Attribute `affinity` was deprecated in version 1.2 and will be removed"
-                " in 1.4. Use `metric` instead",
-                DeprecationWarning,
-            )
-        elif self._metric == "deprecated":
-            self._metric = "euclidean"
-
-        if self._metric == "precomputed":
+        if self.affinity == "precomputed":
             accept_sparse = False
         else:
             accept_sparse = "csr"
         X = self._validate_data(X, accept_sparse=accept_sparse)
-        if self._metric == "precomputed":
+        if self.affinity == "precomputed":
             self.affinity_matrix_ = X
-        elif self._metric == "euclidean":
+        elif self.affinity == "euclidean":
             self.affinity_matrix_ = -euclidean_distances(X, squared=True)
         else:
             raise ValueError(
-                "metric must be 'precomputed' or 'euclidean'. Got %s instead"
-                % str(self._metric)
+                "Affinity must be 'precomputed' or 'euclidean'. Got %s instead"
+                % str(self.affinity)
             )
 
         check_scalar(
@@ -513,7 +496,7 @@ def fit(self, X, y=None):
             random_state=self.random_state,
         )
 
-        if self.metric != "precomputed":
+        if self.affinity != "precomputed":
             self.cluster_centers_ = X[self.cluster_centers_indices_].copy()
 
         return self
@@ -536,7 +519,7 @@ def predict(self, X):
         X = self._validate_data(X, reset=False, accept_sparse="csr")
         if not hasattr(self, "cluster_centers_"):
             raise ValueError(
-                "Predict method is not supported when metric='precomputed'."
+                "Predict method is not supported when affinity='precomputed'."
             )
 
         if self.cluster_centers_.shape[0] > 0:
@@ -552,14 +535,14 @@ def predict(self, X):
             return np.array([-1] * X.shape[0])
 
     def fit_predict(self, X, y=None):
-        """Fit clustering from features/distance matrix; return cluster labels.
+        """Fit clustering from features/affinity matrix; return cluster labels.
 
         Parameters
         ----------
         X : {array-like, sparse matrix} of shape (n_samples, n_features), or \
                 array-like of shape (n_samples, n_samples)
             Training instances to cluster, or similarities / affinities between
-            instances if ``metric='precomputed'``. If a sparse feature matrix
+            instances if ``affinity='precomputed'``. If a sparse feature matrix
             is provided, it will be converted into a sparse ``csr_matrix``.
 
         y : Ignored
diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index b399f805a9d40..9c240a8fcd5ed 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -755,6 +755,10 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
         If "precomputed", a distance matrix (instead of a similarity matrix)
         is needed as input for the fit method.
 
+    .. deprecated:: 1.2
+       `affinity` was deprecated in version 1.2 and will be renamed to `metric`
+        in 1.4.
+
     memory : str or object with the joblib.Memory interface, default=None
         Used to cache the output of the computation of the tree.
         By default, no caching is done. If a string is given, it is the
@@ -876,7 +880,8 @@ def __init__(
         self,
         n_clusters=2,
         *,
-        affinity="euclidean",
+        affinity="deprecated",  # TODO(1.4): Remove
+        metric=None,  # TODO(1.4): Set default as "euclidean"
         memory=None,
         connectivity=None,
         compute_full_tree="auto",
@@ -891,6 +896,7 @@ def __init__(
         self.compute_full_tree = compute_full_tree
         self.linkage = linkage
         self.affinity = affinity
+        self.metric = metric
         self.compute_distances = compute_distances
 
     def fit(self, X, y=None):
@@ -901,7 +907,7 @@ def fit(self, X, y=None):
         X : array-like, shape (n_samples, n_features) or \
                 (n_samples, n_samples)
             Training instances to cluster, or distances between instances if
-            ``affinity='precomputed'``.
+            ``metric='precomputed'``.
 
         y : Ignored
             Not used, present here for API consistency by convention.
@@ -930,6 +936,17 @@ def _fit(self, X):
         """
         memory = check_memory(self.memory)
 
+        # TODO(1.4): Remove
+        self._metric = self.metric or self.affinity
+        if self.affinity != "deprecated":
+            warnings.warn(
+                "Attribute `affinity` was deprecated in version 1.2 and will be removed"
+                " in 1.4. Use `metric` instead",
+                DeprecationWarning,
+            )
+        elif self._metric == "deprecated":
+            self._metric = "euclidean"
+
         if self.n_clusters is not None and self.n_clusters <= 0:
             raise ValueError(
                 "n_clusters should be an integer greater than 0. %s was provided."
@@ -948,10 +965,10 @@ def _fit(self, X):
                 "compute_full_tree must be True if distance_threshold is set."
             )
 
-        if self.linkage == "ward" and self.affinity != "euclidean":
+        if self.linkage == "ward" and self._metric != "euclidean":
             raise ValueError(
                 "%s was provided as affinity. Ward can only "
-                "work with euclidean distances." % (self.affinity,)
+                "work with euclidean distances." % (self._metric,)
             )
 
         if self.linkage not in _TREE_BUILDERS:
@@ -989,7 +1006,7 @@ def _fit(self, X):
         kwargs = {}
         if self.linkage != "ward":
             kwargs["linkage"] = self.linkage
-            kwargs["affinity"] = self.affinity
+            kwargs["affinity"] = self._metric
 
         distance_threshold = self.distance_threshold
 
@@ -1071,6 +1088,10 @@ class FeatureAgglomeration(
         "manhattan", "cosine", or 'precomputed'.
         If linkage is "ward", only "euclidean" is accepted.
 
+    .. deprecated:: 1.2
+       `affinity` was deprecated in version 1.2 and will be renamed to `metric`
+        in 1.4.
+
     memory : str or object with the joblib.Memory interface, default=None
         Used to cache the output of the computation of the tree.
         By default, no caching is done. If a string is given, it is the
@@ -1196,7 +1217,8 @@ def __init__(
         self,
         n_clusters=2,
         *,
-        affinity="euclidean",
+        affinity="deprecated",  # TODO(1.4): Remove
+        metric=None,  # TODO(1.4): Set default as "euclidean"
         memory=None,
         connectivity=None,
         compute_full_tree="auto",
@@ -1212,6 +1234,7 @@ def __init__(
             compute_full_tree=compute_full_tree,
             linkage=linkage,
             affinity=affinity,
+            metric=metric,
             distance_threshold=distance_threshold,
             compute_distances=compute_distances,
         )
diff --git a/sklearn/cluster/tests/test_affinity_propagation.py b/sklearn/cluster/tests/test_affinity_propagation.py
index 57012ab5b69b2..e5dc5d584266d 100644
--- a/sklearn/cluster/tests/test_affinity_propagation.py
+++ b/sklearn/cluster/tests/test_affinity_propagation.py
@@ -45,7 +45,7 @@ def test_affinity_propagation():
     assert n_clusters == n_clusters_
 
     af = AffinityPropagation(
-        preference=preference, metric="precomputed", random_state=28
+        preference=preference, affinity="precomputed", random_state=28
     )
     labels_precomputed = af.fit(S).labels_
 
@@ -82,11 +82,10 @@ def test_affinity_propagation_affinity_shape():
         (X, {"damping": 2}, ValueError, "damping == 2, must be < 1"),
         (X, {"max_iter": 0}, ValueError, "max_iter == 0, must be >= 1."),
         (X, {"convergence_iter": 0}, ValueError, "convergence_iter == 0, must be >= 1"),
-        # TODO(1.4): Replace message "Affinity must be" with "metric must be"
-        (X, {"metric": "unknown"}, ValueError, "metric must be"),
+        (X, {"affinity": "unknown"}, ValueError, "Affinity must be"),
         (
             csr_matrix((3, 3)),
-            {"metric": "precomputed"},
+            {"affinity": "precomputed"},
             TypeError,
             "A sparse matrix was passed, but dense data is required",
         ),
@@ -100,7 +99,7 @@ def test_affinity_propagation_params_validation(input, params, err_type, err_msg
 
 def test_affinity_propagation_predict():
     # Test AffinityPropagation.predict
-    af = AffinityPropagation(metric="euclidean", random_state=63)
+    af = AffinityPropagation(affinity="euclidean", random_state=63)
     labels = af.fit_predict(X)
     labels2 = af.predict(X)
     assert_array_equal(labels, labels2)
@@ -109,13 +108,13 @@ def test_affinity_propagation_predict():
 def test_affinity_propagation_predict_error():
     # Test exception in AffinityPropagation.predict
     # Not fitted.
-    af = AffinityPropagation(metric="euclidean")
+    af = AffinityPropagation(affinity="euclidean")
     with pytest.raises(ValueError):
         af.predict(X)
 
     # Predict not supported when affinity="precomputed".
     S = np.dot(X, X.T)
-    af = AffinityPropagation(metric="precomputed", random_state=57)
+    af = AffinityPropagation(affinity="precomputed", random_state=57)
     af.fit(S)
     with pytest.raises(ValueError):
         af.predict(X)
@@ -187,7 +186,7 @@ def test_affinity_propagation_predict_non_convergence():
 
 def test_affinity_propagation_non_convergence_regressiontest():
     X = np.array([[1, 0, 0, 0, 0, 0], [0, 1, 1, 1, 0, 0], [0, 0, 1, 0, 0, 1]])
-    af = AffinityPropagation(metric="euclidean", max_iter=2, random_state=34)
+    af = AffinityPropagation(affinity="euclidean", max_iter=2, random_state=34)
     msg = (
         "Affinity propagation did not converge, this model may return degenerate"
         " cluster centers and labels."
@@ -259,7 +258,9 @@ def test_affinity_propagation_float32():
     X = np.array(
         [[1, 0, 0, 0], [0, 1, 1, 0], [0, 1, 1, 0], [0, 0, 0, 1]], dtype="float32"
     )
-    afp = AffinityPropagation(preference=1, metric="precomputed", random_state=0).fit(X)
+    afp = AffinityPropagation(preference=1, affinity="precomputed", random_state=0).fit(
+        X
+    )
     expected = np.array([0, 1, 1, 2])
     assert_array_equal(afp.labels_, expected)
 
@@ -267,7 +268,7 @@ def test_affinity_propagation_float32():
 def test_sparse_input_for_predict():
     # Test to make sure sparse inputs are accepted for predict
     # (non-regression test for issue #20049)
-    af = AffinityPropagation(metric="euclidean", random_state=42)
+    af = AffinityPropagation(affinity="euclidean", random_state=42)
     af.fit(X)
     labels = af.predict(csr_matrix((2, 2)))
     assert_array_equal(labels, (2, 2))
@@ -276,32 +277,8 @@ def test_sparse_input_for_predict():
 def test_sparse_input_for_fit_predict():
     # Test to make sure sparse inputs are accepted for fit_predict
     # (non-regression test for issue #20049)
-    af = AffinityPropagation(metric="euclidean", random_state=42)
+    af = AffinityPropagation(affinity="euclidean", random_state=42)
     rng = np.random.RandomState(42)
     X = csr_matrix(rng.randint(0, 2, size=(5, 5)))
     labels = af.fit_predict(X)
     assert_array_equal(labels, (0, 1, 1, 2, 3))
-
-
-# TODO(1.4): Remove
-def test_deprecate_affinity():
-    est = [
-        AffinityPropagation(random_state=42),
-        AffinityPropagation(metric="euclidean", random_state=42),
-    ]
-    with warnings.catch_warnings():
-        warnings.simplefilter("error", DeprecationWarning)
-        for af in est:
-            af.fit(X)
-            af.predict(X)
-            af.fit_predict(X)
-
-    af = AffinityPropagation(affinity="euclidean", random_state=42)
-    msg = (
-        "Attribute `affinity` was deprecated in version 1.2 and will be removed in 1.4."
-        " Use `metric` instead"
-    )
-    with pytest.warns(DeprecationWarning, match=msg):
-        af.fit(X)
-        af.predict(X)
-        af.fit_predict(X)
diff --git a/sklearn/cluster/tests/test_hierarchical.py b/sklearn/cluster/tests/test_hierarchical.py
index c6607779f80fc..b49d1adc9a4b5 100644
--- a/sklearn/cluster/tests/test_hierarchical.py
+++ b/sklearn/cluster/tests/test_hierarchical.py
@@ -8,6 +8,7 @@
 import itertools
 from tempfile import mkdtemp
 import shutil
+import warnings
 import pytest
 from functools import partial
 
@@ -243,7 +244,7 @@ def test_agglomerative_clustering():
     clustering = AgglomerativeClustering(
         n_clusters=10,
         connectivity=connectivity.toarray(),
-        affinity="manhattan",
+        metric="manhattan",
         linkage="ward",
     )
     with pytest.raises(ValueError):
@@ -255,12 +256,12 @@ def test_agglomerative_clustering():
         clustering = AgglomerativeClustering(
             n_clusters=10,
             connectivity=np.ones((n_samples, n_samples)),
-            affinity=affinity,
+            metric=affinity,
             linkage="complete",
         )
         clustering.fit(X)
         clustering2 = AgglomerativeClustering(
-            n_clusters=10, connectivity=None, affinity=affinity, linkage="complete"
+            n_clusters=10, connectivity=None, metric=affinity, linkage="complete"
         )
         clustering2.fit(X)
         assert_almost_equal(
@@ -277,7 +278,7 @@ def test_agglomerative_clustering():
     clustering2 = AgglomerativeClustering(
         n_clusters=10,
         connectivity=connectivity,
-        affinity="precomputed",
+        metric="precomputed",
         linkage="complete",
     )
     clustering2.fit(X_dist)
@@ -291,7 +292,7 @@ def test_agglomerative_clustering_memory_mapped():
     """
     rng = np.random.RandomState(0)
     Xmm = create_memmap_backed_data(rng.randn(50, 100))
-    AgglomerativeClustering(affinity="euclidean", linkage="single").fit(Xmm)
+    AgglomerativeClustering(metric="euclidean", linkage="single").fit(Xmm)
 
 
 def test_ward_agglomeration():
@@ -876,7 +877,7 @@ def test_invalid_shape_precomputed_dist_matrix():
         ValueError,
         match=r"Distance matrix should be square, got matrix of shape \(5, 3\)",
     ):
-        AgglomerativeClustering(affinity="precomputed", linkage="complete").fit(X)
+        AgglomerativeClustering(metric="precomputed", linkage="complete").fit(X)
 
 
 def test_precomputed_connectivity_affinity_with_2_connected_components():
@@ -916,3 +917,27 @@ def test_precomputed_connectivity_affinity_with_2_connected_components():
 
     assert_array_equal(clusterer.labels_, clusterer_precomputed.labels_)
     assert_array_equal(clusterer.children_, clusterer_precomputed.children_)
+
+
+# TODO(1.4): Remove
+def test_deprecate_affinity():
+    rng = np.random.RandomState(42)
+    X = rng.randn(50, 10)
+    est = [
+        AgglomerativeClustering(),
+        AgglomerativeClustering(metric="euclidean"),
+    ]
+    with warnings.catch_warnings():
+        warnings.simplefilter("error", DeprecationWarning)
+        for af in est:
+            af.fit(X)
+            af.fit_predict(X)
+
+    af = AgglomerativeClustering(affinity="euclidean")
+    msg = (
+        "Attribute `affinity` was deprecated in version 1.2 and will be removed in 1.4."
+        " Use `metric` instead"
+    )
+    with pytest.warns(DeprecationWarning, match=msg):
+        af.fit(X)
+        af.fit_predict(X)

From 07bb1d973922bb0a37c542945f82c0cf4a479c3e Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Thu, 26 May 2022 13:23:00 -0400
Subject: [PATCH 04/15] Added changelog entry

---
 doc/whats_new/v1.2.rst | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/doc/whats_new/v1.2.rst b/doc/whats_new/v1.2.rst
index b4804f3c9c8b9..4d94d496acc7c 100644
--- a/doc/whats_new/v1.2.rst
+++ b/doc/whats_new/v1.2.rst
@@ -59,6 +59,10 @@ Changelog
   and both will have their defaults changed to `n_init='auto'` in 1.4.
   :pr:`23038` by :user:`Meekail Zain <micky774>`.
 
+- |API| The `affinity` attribute is now deprecated for
+  :class:`cluster.AgglomerativeClustering` and will be renamed to `metric` in v1.4.
+  :pr:`23470` by :user:`Meekail Zain <micky774>`.
+
 :mod:`sklearn.datasets`
 .......................
 

From 8ec54999c494be7487f4fd2acc7adbaa3093ad37 Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Fri, 27 May 2022 09:58:01 -0400
Subject: [PATCH 05/15] Fixed docstring

---
 sklearn/cluster/_agglomerative.py | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index 9c240a8fcd5ed..32752b09c638e 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -755,9 +755,16 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
         If "precomputed", a distance matrix (instead of a similarity matrix)
         is needed as input for the fit method.
 
-    .. deprecated:: 1.2
-       `affinity` was deprecated in version 1.2 and will be renamed to `metric`
-        in 1.4.
+        .. deprecated:: 1.2
+        `affinity` was deprecated in version 1.2 and will be renamed to `metric`
+            in 1.4.
+
+    metric : str or callable, default='euclidean'
+        Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
+        "manhattan", "cosine", or "precomputed".
+        If linkage is "ward", only "euclidean" is accepted.
+        If "precomputed", a distance matrix (instead of a similarity matrix)
+        is needed as input for the fit method.
 
     memory : str or object with the joblib.Memory interface, default=None
         Used to cache the output of the computation of the tree.
@@ -1088,9 +1095,14 @@ class FeatureAgglomeration(
         "manhattan", "cosine", or 'precomputed'.
         If linkage is "ward", only "euclidean" is accepted.
 
-    .. deprecated:: 1.2
-       `affinity` was deprecated in version 1.2 and will be renamed to `metric`
-        in 1.4.
+        .. deprecated:: 1.2
+        `affinity` was deprecated in version 1.2 and will be renamed to `metric`
+            in 1.4.
+
+    metric : str or callable, default='euclidean'
+        Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
+        "manhattan", "cosine", or 'precomputed'.
+        If linkage is "ward", only "euclidean" is accepted.
 
     memory : str or object with the joblib.Memory interface, default=None
         Used to cache the output of the computation of the tree.

From b6247254d365fc75c59b304cf24525b26c0df8e6 Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Fri, 27 May 2022 10:02:29 -0400
Subject: [PATCH 06/15] Corrected default param and validation

---
 sklearn/cluster/_agglomerative.py | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index 32752b09c638e..c29a74237f532 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -888,7 +888,7 @@ def __init__(
         n_clusters=2,
         *,
         affinity="deprecated",  # TODO(1.4): Remove
-        metric=None,  # TODO(1.4): Set default as "euclidean"
+        metric="euclidean",
         memory=None,
         connectivity=None,
         compute_full_tree="auto",
@@ -943,16 +943,15 @@ def _fit(self, X):
         """
         memory = check_memory(self.memory)
 
+        self._metric = self.metric
         # TODO(1.4): Remove
-        self._metric = self.metric or self.affinity
         if self.affinity != "deprecated":
             warnings.warn(
                 "Attribute `affinity` was deprecated in version 1.2 and will be removed"
                 " in 1.4. Use `metric` instead",
                 DeprecationWarning,
             )
-        elif self._metric == "deprecated":
-            self._metric = "euclidean"
+            self._metric = self.affinity
 
         if self.n_clusters is not None and self.n_clusters <= 0:
             raise ValueError(
@@ -1230,7 +1229,7 @@ def __init__(
         n_clusters=2,
         *,
         affinity="deprecated",  # TODO(1.4): Remove
-        metric=None,  # TODO(1.4): Set default as "euclidean"
+        metric="euclidean",
         memory=None,
         connectivity=None,
         compute_full_tree="auto",

From 27773357320f9c370c5efb20455678ad0d6bfccd Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Fri, 27 May 2022 11:20:36 -0400
Subject: [PATCH 07/15] Fixed spacing in docstring

---
 sklearn/cluster/_agglomerative.py | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index c29a74237f532..d46f85ae8def9 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -756,8 +756,8 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
         is needed as input for the fit method.
 
         .. deprecated:: 1.2
-        `affinity` was deprecated in version 1.2 and will be renamed to `metric`
-            in 1.4.
+            `affinity` was deprecated in version 1.2 and will be renamed to
+            `metric` in 1.4.
 
     metric : str or callable, default='euclidean'
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
@@ -1095,8 +1095,8 @@ class FeatureAgglomeration(
         If linkage is "ward", only "euclidean" is accepted.
 
         .. deprecated:: 1.2
-        `affinity` was deprecated in version 1.2 and will be renamed to `metric`
-            in 1.4.
+            `affinity` was deprecated in version 1.2 and will be renamed to
+            `metric` in 1.4.
 
     metric : str or callable, default='euclidean'
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",

From 2242749e6b37c850b6768587f05a692016d919fb Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Wed, 1 Jun 2022 21:21:23 -0400
Subject: [PATCH 08/15] Addressed reviewer feedback

---
 sklearn/cluster/_agglomerative.py | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index d46f85ae8def9..c527c6ce37e5b 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -766,6 +766,8 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
         If "precomputed", a distance matrix (instead of a similarity matrix)
         is needed as input for the fit method.
 
+        .. versionadded:: 1.2
+
     memory : str or object with the joblib.Memory interface, default=None
         Used to cache the output of the computation of the tree.
         By default, no caching is done. If a string is given, it is the
@@ -973,7 +975,7 @@ def _fit(self, X):
 
         if self.linkage == "ward" and self._metric != "euclidean":
             raise ValueError(
-                "%s was provided as affinity. Ward can only "
+                "%s was provided as metric. Ward can only "
                 "work with euclidean distances." % (self._metric,)
             )
 
@@ -1103,6 +1105,8 @@ class FeatureAgglomeration(
         "manhattan", "cosine", or 'precomputed'.
         If linkage is "ward", only "euclidean" is accepted.
 
+        .. versionadded:: 1.2
+
     memory : str or object with the joblib.Memory interface, default=None
         Used to cache the output of the computation of the tree.
         By default, no caching is done. If a string is given, it is the

From 14196dcac1f811651224b0bae698816f2d833f7e Mon Sep 17 00:00:00 2001
From: Meekail Zain <34613774+Micky774@users.noreply.github.com>
Date: Mon, 6 Jun 2022 13:06:47 -0400
Subject: [PATCH 09/15] Update sklearn/cluster/_agglomerative.py

Co-authored-by: Thomas J. Fan <thomasjpfan@gmail.com>
---
 sklearn/cluster/_agglomerative.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index c527c6ce37e5b..b84b6db039861 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -951,7 +951,7 @@ def _fit(self, X):
             warnings.warn(
                 "Attribute `affinity` was deprecated in version 1.2 and will be removed"
                 " in 1.4. Use `metric` instead",
-                DeprecationWarning,
+                FutureWarning,
             )
             self._metric = self.affinity
 

From b9986d888f6374b80c3eadae049418e10aae89d6 Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Mon, 6 Jun 2022 13:16:25 -0400
Subject: [PATCH 10/15] Updated test

---
 sklearn/cluster/tests/test_hierarchical.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/sklearn/cluster/tests/test_hierarchical.py b/sklearn/cluster/tests/test_hierarchical.py
index b49d1adc9a4b5..23eba480eb879 100644
--- a/sklearn/cluster/tests/test_hierarchical.py
+++ b/sklearn/cluster/tests/test_hierarchical.py
@@ -928,7 +928,7 @@ def test_deprecate_affinity():
         AgglomerativeClustering(metric="euclidean"),
     ]
     with warnings.catch_warnings():
-        warnings.simplefilter("error", DeprecationWarning)
+        warnings.simplefilter("error", FutureWarning)
         for af in est:
             af.fit(X)
             af.fit_predict(X)
@@ -938,6 +938,6 @@ def test_deprecate_affinity():
         "Attribute `affinity` was deprecated in version 1.2 and will be removed in 1.4."
         " Use `metric` instead"
     )
-    with pytest.warns(DeprecationWarning, match=msg):
+    with pytest.warns(FutureWarning, match=msg):
         af.fit(X)
         af.fit_predict(X)

From f5ffe9042872bf4f62047c939c1885ad3e88bc6b Mon Sep 17 00:00:00 2001
From: Micky774 <zainmeekail@gmail.com>
Date: Thu, 9 Jun 2022 12:43:37 -0400
Subject: [PATCH 11/15] Added error handling and testing for edge-case

---
 sklearn/cluster/_agglomerative.py          | 12 ++++++++++--
 sklearn/cluster/tests/test_hierarchical.py |  8 ++++++++
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index b84b6db039861..8488355cc4e3d 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -890,7 +890,7 @@ def __init__(
         n_clusters=2,
         *,
         affinity="deprecated",  # TODO(1.4): Remove
-        metric="euclidean",
+        metric=None,  # TODO(1.4): Set to "euclidean"
         memory=None,
         connectivity=None,
         compute_full_tree="auto",
@@ -948,12 +948,20 @@ def _fit(self, X):
         self._metric = self.metric
         # TODO(1.4): Remove
         if self.affinity != "deprecated":
+            if self.metric is not None:
+                raise ValueError(
+                    "Both `affinity` and `metric` attributes were set. Attribute"
+                    " `affinity` was deprecated in version 1.2 and will be removed in"
+                    " 1.4. To avoid this error, only set the `metric` attribute."
+                )
             warnings.warn(
                 "Attribute `affinity` was deprecated in version 1.2 and will be removed"
                 " in 1.4. Use `metric` instead",
                 FutureWarning,
             )
             self._metric = self.affinity
+        elif self.metric is None:
+            self._metric = "euclidean"
 
         if self.n_clusters is not None and self.n_clusters <= 0:
             raise ValueError(
@@ -1233,7 +1241,7 @@ def __init__(
         n_clusters=2,
         *,
         affinity="deprecated",  # TODO(1.4): Remove
-        metric="euclidean",
+        metric=None,  # TODO(1.4): Set to "euclidean"
         memory=None,
         connectivity=None,
         compute_full_tree="auto",
diff --git a/sklearn/cluster/tests/test_hierarchical.py b/sklearn/cluster/tests/test_hierarchical.py
index 23eba480eb879..32a3e0070bcfe 100644
--- a/sklearn/cluster/tests/test_hierarchical.py
+++ b/sklearn/cluster/tests/test_hierarchical.py
@@ -940,4 +940,12 @@ def test_deprecate_affinity():
     )
     with pytest.warns(FutureWarning, match=msg):
         af.fit(X)
+    with pytest.warns(FutureWarning, match=msg):
+        af.fit_predict(X)
+
+    af = AgglomerativeClustering(metric="euclidean", affinity="euclidean")
+    msg = "Both `affinity` and `metric` attributes were set. Attribute"
+    with pytest.raises(ValueError, match=msg):
+        af.fit(X)
+    with pytest.raises(ValueError, match=msg):
         af.fit_predict(X)

From dbcb021b2e5d26f8534fa016cfbad687da7397d8 Mon Sep 17 00:00:00 2001
From: Meekail Zain <34613774+Micky774@users.noreply.github.com>
Date: Thu, 30 Jun 2022 16:32:09 -0400
Subject: [PATCH 12/15] Apply suggestions from code review

Co-authored-by: Guillaume Lemaitre <g.lemaitre58@gmail.com>
---
 sklearn/cluster/_agglomerative.py | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index 02fdf30a7d7e6..1932231593505 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -764,7 +764,7 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
             `affinity` was deprecated in version 1.2 and will be renamed to
             `metric` in 1.4.
 
-    metric : str or callable, default='euclidean'
+    metric : str or callable, default=None
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
         "manhattan", "cosine", or "precomputed".
         If linkage is "ward", only "euclidean" is accepted.
@@ -989,11 +989,6 @@ def _fit(self, X):
         elif self.metric is None:
             self._metric = "euclidean"
 
-        if self.n_clusters is not None and self.n_clusters <= 0:
-            raise ValueError(
-                "n_clusters should be an integer greater than 0. %s was provided."
-                % str(self.n_clusters)
-            )
 
         if not ((self.n_clusters is None) ^ (self.distance_threshold is None)):
             raise ValueError(
@@ -1009,8 +1004,8 @@ def _fit(self, X):
 
         if self.linkage == "ward" and self._metric != "euclidean":
             raise ValueError(
-                "%s was provided as metric. Ward can only "
-                "work with euclidean distances." % (self._metric,)
+                f"{self._metric} was provided as metric. Ward can only "
+                "work with euclidean distances."
             )
 
         tree_builder = _TREE_BUILDERS[self.linkage]

From 294946acfd86e9fa575d5bd08e5e4aed8558ab40 Mon Sep 17 00:00:00 2001
From: Meekail Zain <zainmeekail@gmail.com>
Date: Thu, 30 Jun 2022 16:35:55 -0400
Subject: [PATCH 13/15] Streamlined test and renamed var

---
 sklearn/cluster/_agglomerative.py          | 17 +++++++++--------
 sklearn/cluster/tests/test_hierarchical.py | 16 +++-------------
 2 files changed, 12 insertions(+), 21 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index 1932231593505..cd869aa44b0e3 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -766,10 +766,10 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
 
     metric : str or callable, default=None
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
-        "manhattan", "cosine", or "precomputed".
-        If linkage is "ward", only "euclidean" is accepted.
-        If "precomputed", a distance matrix (instead of a similarity matrix)
-        is needed as input for the fit method.
+        "manhattan", "cosine", or "precomputed". If set to `None` then
+        "euclidean" is used. If linkage is "ward", only "euclidean" is
+        accepted. If "precomputed", a distance matrix (instead of a similarity
+        matrix) is needed as input for the fit method.
 
         .. versionadded:: 1.2
 
@@ -989,7 +989,6 @@ def _fit(self, X):
         elif self.metric is None:
             self._metric = "euclidean"
 
-
         if not ((self.n_clusters is None) ^ (self.distance_threshold is None)):
             raise ValueError(
                 "Exactly one of n_clusters and "
@@ -1128,10 +1127,12 @@ class FeatureAgglomeration(
             `affinity` was deprecated in version 1.2 and will be renamed to
             `metric` in 1.4.
 
-    metric : str or callable, default='euclidean'
+    metric : str or callable, default=None
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
-        "manhattan", "cosine", or 'precomputed'.
-        If linkage is "ward", only "euclidean" is accepted.
+        "manhattan", "cosine", or "precomputed". If set to `None` then
+        "euclidean" is used. If linkage is "ward", only "euclidean" is
+        accepted. If "precomputed", a distance matrix (instead of a similarity
+        matrix) is needed as input for the fit method.
 
         .. versionadded:: 1.2
 
diff --git a/sklearn/cluster/tests/test_hierarchical.py b/sklearn/cluster/tests/test_hierarchical.py
index 01cac453e5b97..012073d1ba332 100644
--- a/sklearn/cluster/tests/test_hierarchical.py
+++ b/sklearn/cluster/tests/test_hierarchical.py
@@ -8,7 +8,6 @@
 import itertools
 from tempfile import mkdtemp
 import shutil
-import warnings
 import pytest
 from functools import partial
 
@@ -249,17 +248,17 @@ def test_agglomerative_clustering():
         clustering.fit(X)
 
     # Test using another metric than euclidean works with linkage complete
-    for affinity in PAIRED_DISTANCES.keys():
+    for metric in PAIRED_DISTANCES.keys():
         # Compare our (structured) implementation to scipy
         clustering = AgglomerativeClustering(
             n_clusters=10,
             connectivity=np.ones((n_samples, n_samples)),
-            metric=affinity,
+            metric=metric,
             linkage="complete",
         )
         clustering.fit(X)
         clustering2 = AgglomerativeClustering(
-            n_clusters=10, connectivity=None, metric=affinity, linkage="complete"
+            n_clusters=10, connectivity=None, metric=metric, linkage="complete"
         )
         clustering2.fit(X)
         assert_almost_equal(
@@ -907,15 +906,6 @@ def test_precomputed_connectivity_affinity_with_2_connected_components():
 def test_deprecate_affinity():
     rng = np.random.RandomState(42)
     X = rng.randn(50, 10)
-    est = [
-        AgglomerativeClustering(),
-        AgglomerativeClustering(metric="euclidean"),
-    ]
-    with warnings.catch_warnings():
-        warnings.simplefilter("error", FutureWarning)
-        for af in est:
-            af.fit(X)
-            af.fit_predict(X)
 
     af = AgglomerativeClustering(affinity="euclidean")
     msg = (

From e62ed6aac96a63484da3d7094e05444f587476ed Mon Sep 17 00:00:00 2001
From: Meekail Zain <zainmeekail@gmail.com>
Date: Mon, 4 Jul 2022 16:41:07 -0400
Subject: [PATCH 14/15] Clarified docstring per review feedback

---
 sklearn/cluster/_agglomerative.py | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/sklearn/cluster/_agglomerative.py b/sklearn/cluster/_agglomerative.py
index cd869aa44b0e3..90ee3336e0478 100644
--- a/sklearn/cluster/_agglomerative.py
+++ b/sklearn/cluster/_agglomerative.py
@@ -768,8 +768,8 @@ class AgglomerativeClustering(ClusterMixin, BaseEstimator):
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
         "manhattan", "cosine", or "precomputed". If set to `None` then
         "euclidean" is used. If linkage is "ward", only "euclidean" is
-        accepted. If "precomputed", a distance matrix (instead of a similarity
-        matrix) is needed as input for the fit method.
+        accepted. If "precomputed", a distance matrix is needed as input for
+        the fit method.
 
         .. versionadded:: 1.2
 
@@ -1131,8 +1131,8 @@ class FeatureAgglomeration(
         Metric used to compute the linkage. Can be "euclidean", "l1", "l2",
         "manhattan", "cosine", or "precomputed". If set to `None` then
         "euclidean" is used. If linkage is "ward", only "euclidean" is
-        accepted. If "precomputed", a distance matrix (instead of a similarity
-        matrix) is needed as input for the fit method.
+        accepted. If "precomputed", a distance matrix is needed as input for
+        the fit method.
 
         .. versionadded:: 1.2
 

From 43ff645f193a33423c7119c297a7b0e7b03c5280 Mon Sep 17 00:00:00 2001
From: Meekail Zain <zainmeekail@gmail.com>
Date: Wed, 6 Jul 2022 09:56:53 -0400
Subject: [PATCH 15/15] Updated affinity-->metric in plotting

---
 examples/cluster/plot_agglomerative_clustering_metrics.py | 4 ++--
 examples/cluster/plot_cluster_comparison.py               | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/examples/cluster/plot_agglomerative_clustering_metrics.py b/examples/cluster/plot_agglomerative_clustering_metrics.py
index e022035ca2207..38fd3682d48ec 100644
--- a/examples/cluster/plot_agglomerative_clustering_metrics.py
+++ b/examples/cluster/plot_agglomerative_clustering_metrics.py
@@ -125,7 +125,7 @@ def sqr(x):
 # Plot clustering results
 for index, metric in enumerate(["cosine", "euclidean", "cityblock"]):
     model = AgglomerativeClustering(
-        n_clusters=n_clusters, linkage="average", affinity=metric
+        n_clusters=n_clusters, linkage="average", metric=metric
     )
     model.fit(X)
     plt.figure()
@@ -134,7 +134,7 @@ def sqr(x):
         plt.plot(X[model.labels_ == l].T, c=c, alpha=0.5)
     plt.axis("tight")
     plt.axis("off")
-    plt.suptitle("AgglomerativeClustering(affinity=%s)" % metric, size=20)
+    plt.suptitle("AgglomerativeClustering(metric=%s)" % metric, size=20)
 
 
 plt.show()
diff --git a/examples/cluster/plot_cluster_comparison.py b/examples/cluster/plot_cluster_comparison.py
index 8b52759c79018..a9e39267411b7 100644
--- a/examples/cluster/plot_cluster_comparison.py
+++ b/examples/cluster/plot_cluster_comparison.py
@@ -171,7 +171,7 @@
     )
     average_linkage = cluster.AgglomerativeClustering(
         linkage="average",
-        affinity="cityblock",
+        metric="cityblock",
         n_clusters=params["n_clusters"],
         connectivity=connectivity,
     )