ENH Add sparse input support to OPTICS (#22965)

Micky774 · huntzhan · elefantinerrrrs · web-flow · commit af5b6a100357 · 2022-05-02T11:37:54.000+02:00
Co-authored-by: huntzhan &lt;huntzhan.dev@gmail.com&gt;
Co-authored-by: Clickedbigfoot &lt;clickedbigfoot@gmail.com&gt;
Co-authored-by: Jérémie du Boisberranger &lt;34657725+jeremiedbb@users.noreply.github.com&gt;
Co-authored-by: Thomas J. Fan &lt;thomasjpfan@gmail.com&gt;
diff --git a/doc/whats_new/v1.2.rst b/doc/whats_new/v1.2.rst
@@ -33,6 +33,14 @@ Changelog
     :pr:`123456` by :user:`Joe Bloggs <joeongithub>`.
     where 123456 is the *pull request* number, not the issue number.
 
+:mod:`sklearn.cluster`
+......................
+
+- |Enhancement| The `predict` and `fit_predict` methods of :class:`cluster.OPTICS` now
+  accept sparse data type for input data. :pr:`14736` by :user:`Hunt Zhan <huntzhan>`,
+  :pr:`20802` by :user:`Brandon Pokorny <Clickedbigfoot>`,
+  and :pr:`22965` by :user:`Meekail Zain <micky774>`.
+
 Code and Documentation Contributors
 -----------------------------------
 
diff --git a/sklearn/cluster/_optics.py b/sklearn/cluster/_optics.py
@@ -20,6 +20,7 @@
 from ..neighbors import NearestNeighbors
 from ..base import BaseEstimator, ClusterMixin
 from ..metrics import pairwise_distances
+from scipy.sparse import issparse, SparseEfficiencyWarning
 
 
 class OPTICS(ClusterMixin, BaseEstimator):
@@ -81,6 +82,7 @@ class OPTICS(ClusterMixin, BaseEstimator):
           'seuclidean', 'sokalmichener', 'sokalsneath', 'sqeuclidean',
           'yule']
 
+        Sparse matrices are only supported by scikit-learn metrics.
         See the documentation for scipy.spatial.distance for details on these
         metrics.
 
@@ -263,10 +265,11 @@ def fit(self, X, y=None):
 
         Parameters
         ----------
-        X : ndarray of shape (n_samples, n_features), or \
+        X : {ndarray, sparse matrix} of shape (n_samples, n_features), or \
                 (n_samples, n_samples) if metric=’precomputed’
             A feature array, or array of distances between samples if
-            metric='precomputed'.
+            metric='precomputed'. If a sparse matrix is provided, it will be
+            converted into CSR format.
 
         y : Ignored
             Not used, present for API consistency by convention.
@@ -285,7 +288,13 @@ def fit(self, X, y=None):
             )
             warnings.warn(msg, DataConversionWarning)
 
-        X = self._validate_data(X, dtype=dtype)
+        X = self._validate_data(X, dtype=dtype, accept_sparse="csr")
+        if self.metric == "precomputed" and issparse(X):
+            with warnings.catch_warnings():
+                warnings.simplefilter("ignore", SparseEfficiencyWarning)
+                # Set each diagonal to an explicit value so each point is its
+                # own neighbor
+                X.setdiag(X.diagonal())
         memory = check_memory(self.memory)
 
         if self.cluster_method not in ["dbscan", "xi"]:
@@ -603,15 +612,16 @@ def _set_reach_dist(
     # Only compute distances to unprocessed neighbors:
     if metric == "precomputed":
         dists = X[point_index, unproc]
+        if issparse(dists):
+            dists.sort_indices()
+            dists = dists.data
     else:
         _params = dict() if metric_params is None else metric_params.copy()
         if metric == "minkowski" and "p" not in _params:
             # the same logic as neighbors, p is ignored if explicitly set
             # in the dict params
             _params["p"] = p
-        dists = pairwise_distances(
-            P, np.take(X, unproc, axis=0), metric=metric, n_jobs=None, **_params
-        ).ravel()
+        dists = pairwise_distances(P, X[unproc], metric, n_jobs=None, **_params).ravel()
 
     rdists = np.maximum(dists, core_distances_[point_index])
     np.around(rdists, decimals=np.finfo(rdists.dtype).precision, out=rdists)
diff --git a/sklearn/cluster/tests/test_optics.py b/sklearn/cluster/tests/test_optics.py
@@ -3,6 +3,7 @@
 # License: BSD 3 clause
 import numpy as np
 import pytest
+from scipy import sparse
 import warnings
 
 from sklearn.datasets import make_blobs
@@ -15,7 +16,7 @@
 from sklearn.utils import shuffle
 from sklearn.utils._testing import assert_array_equal
 from sklearn.utils._testing import assert_allclose
-
+from sklearn.exceptions import EfficiencyWarning
 from sklearn.cluster.tests.common import generate_clustered_data
 
 
@@ -158,15 +159,19 @@ def test_cluster_hierarchy_(global_dtype):
     assert diff / len(X) < 0.05
 
 
-def test_correct_number_of_clusters():
+@pytest.mark.parametrize(
+    "metric, is_sparse",
+    [["minkowski", False], ["euclidean", True]],
+)
+def test_correct_number_of_clusters(metric, is_sparse):
     # in 'auto' mode
 
     n_clusters = 3
     X = generate_clustered_data(n_clusters=n_clusters)
     # Parameters chosen specifically for this task.
     # Compute OPTICS
-    clust = OPTICS(max_eps=5.0 * 6.0, min_samples=4, xi=0.1)
-    clust.fit(X)
+    clust = OPTICS(max_eps=5.0 * 6.0, min_samples=4, xi=0.1, metric=metric)
+    clust.fit(sparse.csr_matrix(X) if is_sparse else X)
     # number of clusters, ignoring noise if present
     n_clusters_1 = len(set(clust.labels_)) - int(-1 in clust.labels_)
     assert n_clusters_1 == n_clusters
@@ -286,18 +291,25 @@ def test_close_extract():
 
 @pytest.mark.parametrize("eps", [0.1, 0.3, 0.5])
 @pytest.mark.parametrize("min_samples", [3, 10, 20])
-def test_dbscan_optics_parity(eps, min_samples, global_dtype):
+@pytest.mark.parametrize(
+    "metric, is_sparse",
+    [["minkowski", False], ["euclidean", False], ["euclidean", True]],
+)
+def test_dbscan_optics_parity(eps, min_samples, metric, is_sparse, global_dtype):
     # Test that OPTICS clustering labels are <= 5% difference of DBSCAN
 
     centers = [[1, 1], [-1, -1], [1, -1]]
     X, labels_true = make_blobs(
         n_samples=750, centers=centers, cluster_std=0.4, random_state=0
     )
+    X = sparse.csr_matrix(X) if is_sparse else X
 
     X = X.astype(global_dtype, copy=False)
 
     # calculate optics with dbscan extract at 0.3 epsilon
-    op = OPTICS(min_samples=min_samples, cluster_method="dbscan", eps=eps).fit(X)
+    op = OPTICS(
+        min_samples=min_samples, cluster_method="dbscan", eps=eps, metric=metric
+    ).fit(X)
 
     # calculate dbscan labels
     db = DBSCAN(eps=eps, min_samples=min_samples).fit(X)
@@ -344,7 +356,8 @@ def test_min_cluster_size(min_cluster_size, global_dtype):
         assert min(cluster_sizes) >= min_cluster_size
     # check behaviour is the same when min_cluster_size is a fraction
     clust_frac = OPTICS(
-        min_samples=9, min_cluster_size=min_cluster_size / redX.shape[0]
+        min_samples=9,
+        min_cluster_size=min_cluster_size / redX.shape[0],
     )
     clust_frac.fit(redX)
     assert_array_equal(clust.labels_, clust_frac.labels_)
@@ -356,17 +369,26 @@ def test_min_cluster_size_invalid(min_cluster_size):
     with pytest.raises(ValueError, match="must be a positive integer or a "):
         clust.fit(X)
 
+    clust = OPTICS(min_cluster_size=min_cluster_size, metric="euclidean")
+    with pytest.raises(ValueError, match="must be a positive integer or a "):
+        clust.fit(sparse.csr_matrix(X))
+
 
 def test_min_cluster_size_invalid2():
     clust = OPTICS(min_cluster_size=len(X) + 1)
     with pytest.raises(ValueError, match="must be no greater than the "):
         clust.fit(X)
 
+    clust = OPTICS(min_cluster_size=len(X) + 1, metric="euclidean")
+    with pytest.raises(ValueError, match="must be no greater than the "):
+        clust.fit(sparse.csr_matrix(X))
+
 
 def test_processing_order():
     # Ensure that we consider all unprocessed points,
     # not only direct neighbors. when picking the next point.
     Y = [[0], [10], [-10], [25]]
+
     clust = OPTICS(min_samples=3, max_eps=15).fit(Y)
     assert_array_equal(clust.reachability_, [np.inf, 10, 10, 15])
     assert_array_equal(clust.core_distances_, [10, 15, np.inf, np.inf])
@@ -796,10 +818,16 @@ def test_extract_dbscan(global_dtype):
     assert_array_equal(np.sort(np.unique(clust.labels_)), [0, 1, 2, 3])
 
 
-def test_precomputed_dists(global_dtype):
+@pytest.mark.parametrize("is_sparse", [False, True])
+def test_precomputed_dists(is_sparse, global_dtype):
     redX = X[::2].astype(global_dtype, copy=False)
     dists = pairwise_distances(redX, metric="euclidean")
-    clust1 = OPTICS(min_samples=10, algorithm="brute", metric="precomputed").fit(dists)
+    dists = sparse.csr_matrix(dists) if is_sparse else dists
+    with warnings.catch_warnings():
+        warnings.simplefilter("ignore", EfficiencyWarning)
+        clust1 = OPTICS(min_samples=10, algorithm="brute", metric="precomputed").fit(
+            dists
+        )
     clust2 = OPTICS(min_samples=10, algorithm="brute", metric="euclidean").fit(redX)
 
     assert_allclose(clust1.reachability_, clust2.reachability_)
