Repairs chunk generation for unknown labels, solves issue #260 (#263)

* chunks return a map of index to chunk

* maj

* maj

* remove storing of known labels

* typo

* no self.num_points

* tests for unlabeled, repairs chunk generation

* maj

* testing diff features

* corrected test

* diff warning

* maj

* added parameter bound test

Author: RobinVogel

metric_learn/constraints.py

@@ -5,7 +5,6 @@
 import numpy as np
 import warnings
 from six.moves import xrange
-from scipy.sparse import coo_matrix
 from sklearn.utils import check_random_state
 
 __all__ = ['Constraints']
@@ -20,21 +19,7 @@ class Constraints(object):
   def __init__(self, partial_labels):
     '''partial_labels : int arraylike, -1 indicating unknown label'''
     partial_labels = np.asanyarray(partial_labels, dtype=int)
-    self.num_points, = partial_labels.shape
-    self.known_label_idx, = np.where(partial_labels >= 0)
-    self.known_labels = partial_labels[self.known_label_idx]
-
-  def adjacency_matrix(self, num_constraints, random_state=None):
-    random_state = check_random_state(random_state)
-    a, b, c, d = self.positive_negative_pairs(num_constraints,
-                                              random_state=random_state)
-    row = np.concatenate((a, c))
-    col = np.concatenate((b, d))
-    data = np.ones_like(row, dtype=int)
-    data[len(a):] = -1
-    adj = coo_matrix((data, (row, col)), shape=(self.num_points,) * 2)
-    # symmetrize
-    return adj + adj.T
+    self.partial_labels = partial_labels
 
   def positive_negative_pairs(self, num_constraints, same_length=False,
                               random_state=None):
@@ -50,17 +35,19 @@ def positive_negative_pairs(self, num_constraints, same_length=False,
 
   def _pairs(self, num_constraints, same_label=True, max_iter=10,
              random_state=np.random):
-    num_labels = len(self.known_labels)
+    known_label_idx, = np.where(self.partial_labels >= 0)
+    known_labels = self.partial_labels[known_label_idx]
+    num_labels = len(known_labels)
     ab = set()
     it = 0
     while it < max_iter and len(ab) < num_constraints:
       nc = num_constraints - len(ab)
       for aidx in random_state.randint(num_labels, size=nc):
         if same_label:
-          mask = self.known_labels[aidx] == self.known_labels
+          mask = known_labels[aidx] == known_labels
           mask[aidx] = False  # avoid identity pairs
         else:
-          mask = self.known_labels[aidx] != self.known_labels
+          mask = known_labels[aidx] != known_labels
         b_choices, = np.where(mask)
         if len(b_choices) > 0:
           ab.add((aidx, random_state.choice(b_choices)))
@@ -69,16 +56,18 @@ def _pairs(self, num_constraints, same_label=True, max_iter=10,
       warnings.warn("Only generated %d %s constraints (requested %d)" % (
           len(ab), 'positive' if same_label else 'negative', num_constraints))
     ab = np.array(list(ab)[:num_constraints], dtype=int)
-    return self.known_label_idx[ab.T]
+    return known_label_idx[ab.T]
 
   def chunks(self, num_chunks=100, chunk_size=2, random_state=None):
     """
     the random state object to be passed must be a numpy random seed
     """
     random_state = check_random_state(random_state)
-    chunks = -np.ones_like(self.known_label_idx, dtype=int)
-    uniq, lookup = np.unique(self.known_labels, return_inverse=True)
-    all_inds = [set(np.where(lookup == c)[0]) for c in xrange(len(uniq))]
+    chunks = -np.ones_like(self.partial_labels, dtype=int)
+    uniq, lookup = np.unique(self.partial_labels, return_inverse=True)
+    unknown_uniq = np.where(uniq < 0)[0]
+    all_inds = [set(np.where(lookup == c)[0]) for c in xrange(len(uniq))
+                if c not in unknown_uniq]
     max_chunks = int(np.sum([len(s) // chunk_size for s in all_inds]))
     if max_chunks < num_chunks:
       raise ValueError(('Not enough possible chunks of %d elements in each'

metric_learn/rca.py

@@ -93,10 +93,12 @@ def __init__(self, n_components=None, num_dims='deprecated',
 
   def _check_dimension(self, rank, X):
     d = X.shape[1]
+
     if rank < d:
       warnings.warn('The inner covariance matrix is not invertible, '
                     'so the transformation matrix may contain Nan values. '
-                    'You should reduce the dimensionality of your input,'
+                    'You should remove any linearly dependent features and/or '
+                    'reduce the dimensionality of your input, '
                     'for instance using `sklearn.decomposition.PCA` as a '
                     'preprocessing step.')
 
@@ -241,4 +243,13 @@ def fit(self, X, y, random_state='deprecated'):
     chunks = Constraints(y).chunks(num_chunks=self.num_chunks,
                                    chunk_size=self.chunk_size,
                                    random_state=self.random_state)
+
+    if self.num_chunks * (self.chunk_size - 1) < X.shape[1]:
+      warnings.warn('Due to the parameters of RCA_Supervised, '
+                    'the inner covariance matrix is not invertible, '
+                    'so the transformation matrix will contain Nan values. '
+                    'Increase the number or size of the chunks to correct '
+                    'this problem.'
+                    )
+
     return RCA.fit(self, X, chunks)

test/metric_learn_test.py

@@ -1100,9 +1100,11 @@ def test_rank_deficient_returns_warning(self):
     rca = RCA()
     msg = ('The inner covariance matrix is not invertible, '
            'so the transformation matrix may contain Nan values. '
-           'You should reduce the dimensionality of your input,'
+           'You should remove any linearly dependent features and/or '
+           'reduce the dimensionality of your input, '
            'for instance using `sklearn.decomposition.PCA` as a '
            'preprocessing step.')
+
     with pytest.warns(None) as raised_warnings:
       rca.fit(X, y)
     assert any(str(w.message) == msg for w in raised_warnings)
@@ -1136,6 +1138,41 @@ def test_changed_behaviour_warning_random_state(self):
       rca_supervised.fit(X, y)
     assert any(msg == str(wrn.message) for wrn in raised_warning)
 
+  def test_unknown_labels(self):
+    n = 200
+    num_chunks = 50
+    X, y = make_classification(random_state=42, n_samples=2 * n,
+                               n_features=6, n_informative=6, n_redundant=0)
+    y2 = np.concatenate((y[:n], -np.ones(n)))
+
+    rca = RCA_Supervised(num_chunks=num_chunks, random_state=42)
+    rca.fit(X[:n], y[:n])
+
+    rca2 = RCA_Supervised(num_chunks=num_chunks, random_state=42)
+    rca2.fit(X, y2)
+
+    assert not np.any(np.isnan(rca.components_))
+    assert not np.any(np.isnan(rca2.components_))
+
+    np.testing.assert_array_equal(rca.components_, rca2.components_)
+
+  def test_bad_parameters(self):
+    n = 200
+    num_chunks = 3
+    X, y = make_classification(random_state=42, n_samples=n,
+                               n_features=6, n_informative=6, n_redundant=0)
+
+    rca = RCA_Supervised(num_chunks=num_chunks, random_state=42)
+    msg = ('Due to the parameters of RCA_Supervised, '
+           'the inner covariance matrix is not invertible, '
+           'so the transformation matrix will contain Nan values. '
+           'Increase the number or size of the chunks to correct '
+           'this problem.'
+           )
+    with pytest.warns(None) as raised_warning:
+      rca.fit(X, y)
+    assert any(str(w.message) == msg for w in raised_warning)
+
 
 @pytest.mark.parametrize('num_dims', [None, 2])
 def test_deprecation_num_dims_rca(num_dims):

test/test_constraints.py

@@ -1,4 +1,3 @@
-import unittest
 import pytest
 import numpy as np
 from sklearn.utils import shuffle
@@ -34,7 +33,8 @@ def test_exact_num_points_for_chunks(num_chunks, chunk_size):
   chunks = constraints.chunks(num_chunks=num_chunks, chunk_size=chunk_size,
                               random_state=SEED)
 
-  chunk_no, size_each_chunk = np.unique(chunks, return_counts=True)
+  chunk_no, size_each_chunk = np.unique(chunks[chunks >= 0],
+                                        return_counts=True)
 
   np.testing.assert_array_equal(size_each_chunk, chunk_size)
   assert chunk_no.shape[0] == num_chunks
@@ -59,5 +59,13 @@ def test_chunk_case_one_miss_point(num_chunks, chunk_size):
   assert str(e.value) == expected_message
 
 
-if __name__ == '__main__':
-  unittest.main()
+@pytest.mark.parametrize("num_chunks, chunk_size", [(5, 10), (10, 50)])
+def test_unknown_labels_not_in_chunks(num_chunks, chunk_size):
+  """Checks that unknown labels are not assigned to any chunk."""
+  labels = gen_labels_for_chunks(num_chunks, chunk_size)
+
+  constraints = Constraints(labels)
+  chunks = constraints.chunks(num_chunks=num_chunks, chunk_size=chunk_size,
+                              random_state=SEED)
+
+  assert np.all(chunks[labels < 0] < 0)