Start on basic tests for robust single linkage

Author: lmcinnes

hdbscan/tests/test_rsl.py

@@ -0,0 +1,61 @@
+"""
+Tests for Robust Single Linkage clustering algorithm
+"""
+#import pickle
+from nose.tools import assert_less
+import numpy as np
+from scipy.spatial import distance
+from scipy import sparse
+from sklearn.utils.testing import assert_equal
+from sklearn.utils.testing import assert_array_equal
+from sklearn.utils.testing import assert_raises
+from sklearn.utils.testing import assert_in
+from sklearn.utils.testing import assert_not_in
+from sklearn.utils.testing import assert_no_warnings
+from sklearn.utils.testing import if_matplotlib
+from hdbscan import RobustSingleLinkage
+from hdbscan import robust_single_linkage
+from sklearn.cluster.tests.common import generate_clustered_data
+
+from sklearn import datasets
+
+n_clusters = 3
+X = generate_clustered_data(n_clusters=n_clusters, n_samples_per_cluster=50)
+
+def test_rsl_distance_matrix():
+    D = distance.squareform(distance.pdist(X))
+    D /= np.max(D)
+
+    labels, tree = robust_single_linkage(D, 0.25, metric='precomputed')
+    # number of clusters, ignoring noise if present
+    n_clusters_1 = len(set(labels)) - int(-1 in labels) # ignore noise
+    #assert_equal(n_clusters_1, n_clusters)
+
+    labels = RobustSingleLinkage(metric="precomputed").fit(D).labels_
+    n_clusters_2 = len(set(labels)) - int(-1 in labels)
+    #assert_equal(n_clusters_2, n_clusters)
+
+def test_rsl_feature_vector():
+    labels, tree = robust_single_linkage(X, 0.2)
+    n_clusters_1 = len(set(labels)) - int(-1 in labels)
+    #assert_equal(n_clusters_1, n_clusters)
+
+    labels = RobustSingleLinkage().fit(X).labels_
+    n_clusters_2 = len(set(labels)) - int(-1 in labels)
+    #assert_equal(n_clusters_2, n_clusters)
+
+def test_rsl_callable_metric():
+    # metric is the function reference, not the string key.
+    metric = distance.euclidean
+
+    labels, tree = robust_single_linkage(X, 0.2, metric=metric)
+    n_clusters_1 = len(set(labels)) - int(-1 in labels)
+    #assert_equal(n_clusters_1, n_clusters)
+
+    labels = RobustSingleLinkage(metric=metric).fit(X).labels_
+    n_clusters_2 = len(set(labels)) - int(-1 in labels)
+    #assert_equal(n_clusters_2, n_clusters)
+
+def test_rsl_input_lists():
+    X = [[1., 2.], [3., 4.]]
+    RobustSingleLinkage().fit(X)  # must not raise exception