Improved test case and removed duplicate test

doc/examples/plot_rag_merge.py

@@ -34,13 +34,12 @@ def _weight_mean_color(graph, src, dst, n):
         The absolute difference of the mean color between node `dst` and `n`.
     """
 
-    #print 'merging
     diff = graph.node[dst]['mean color'] - graph.node[n]['mean color']
     diff = np.linalg.norm(diff)
     return diff
 
 
-def _pre_merge_mean_color(graph, src, dst):
+def merge_mean_color(graph, src, dst):
     """Callback called before merging two nodes of a mean color distance graph.
 
     This method computes the mean color of `dst`.
@@ -58,44 +57,13 @@ def _pre_merge_mean_color(graph, src, dst):
                                      graph.node[dst]['pixel count'])
 
 
-def merge_hierarchical_mean_color(labels, rag, thresh, rag_copy=True,
-                                  in_place_merge=False):
-    """Perform hierarchical merging of a color distance RAG.
-
-    Greedily merges the most similar pair of nodes until no edges lower than
-    `thresh` remain.
-
-    Parameters
-    ----------
-    labels : ndarray
-        The array of labels.
-    rag : RAG
-        The Region Adjacency Graph.
-    thresh : float
-        Regions connected by an edge with weight smaller than `thresh` are
-        merged.
-    rag_copy : bool, optional
-        If set, the RAG copied before modifying.
-    in_place_merge : bool, optional
-        If set, the nodes are merged in place. Otherwise, a new node is
-        created for each merge.
-
-    Examples
-    --------
-    >>> from skimage import data, graph, segmentation
-    >>> img = data.coffee()
-    >>> labels = segmentation.slic(img)
-    >>> rag = graph.rag_mean_color(img, labels)
-    >>> new_labels = graph.merge_hierarchical_mean_color(labels, rag, 40)
-    """
-    return graph.merge_hierarchical(labels, rag, thresh, rag_copy,
-                                    in_place_merge, _pre_merge_mean_color,
-                                    _weight_mean_color)
-
 img = data.coffee()
 labels = segmentation.slic(img, compactness=30, n_segments=400)
 g = graph.rag_mean_color(img, labels)
-labels2 = merge_hierarchical_mean_color(labels, g, 40)
+
+labels2 = graph.merge_hierarchical(labels, g, 40, False, True,
+                                   merge_mean_color, _weight_mean_color)
+
 g2 = graph.rag_mean_color(img, labels2)
 
 out = color.label2rgb(labels2, img, kind='avg')

skimage/graph/tests/test_rag.py

@@ -2,7 +2,7 @@ import numpy as np
 from skimage import graph
 from skimage._shared.version_requirements import is_installed
 from numpy.testing.decorators import skipif
-from skimage import segmentation, io
+from skimage import segmentation
 from numpy import testing
 
 
@@ -110,17 +110,7 @@ def test_rag_error():
                           2, 'non existant mode')
 
 
-@skipif(not is_installed('networkx'))
-def test_rag_error():
-    img = np.zeros((10, 10, 3), dtype='uint8')
-    labels = np.zeros((10, 10), dtype='uint8')
-    labels[:5, :] = 0
-    labels[5:, :] = 1
-    testing.assert_raises(ValueError, graph.rag_mean_color, img, labels,
-                          2, 'non existant mode')
-
 def _weight_mean_color(graph, src, dst, n):
-    #print 'merging
     diff = graph.node[dst]['mean color'] - graph.node[n]['mean color']
     diff = np.linalg.norm(diff)
     return diff
@@ -139,20 +129,31 @@ def merge_hierarchical_mean_color(labels, rag, thresh, rag_copy=True,
                                     in_place_merge, _pre_merge_mean_color,
                                     _weight_mean_color)
 
+
 @skipif(not is_installed('networkx'))
 def test_rag_hierarchical():
     img = np.zeros((8, 8, 3), dtype='uint8')
     labels = np.zeros((8, 8), dtype='uint8')
 
-    img[:, :, :] = 128
-    labels[:,:] = 1
+    img[:, :, :] = 30
+    labels[:, :] = 1
 
-    img[0:4,0:4,:] = 255,255,255
+    img[0:4, 0:4, :] = 10, 10, 10
     labels[0:4, 0:4] = 2
 
-    img[4:, 0:4,:] = 0,0,0
+    img[4:, 0:4, :] = 20, 20, 20
     labels[4:, 0:4] = 3
 
     g = graph.rag_mean_color(img, labels)
-    result = merge_hierarchical_mean_color(labels, g, 300)
+    g2 = g.copy()
+    thresh = 17.3206  # just above 10*sqrt(3)
+
+    result = merge_hierarchical_mean_color(labels, g, thresh)
+    assert len(np.unique(result)) == 2
+
+    result = merge_hierarchical_mean_color(labels, g2, thresh,
+                                           in_place_merge=True)
+    assert len(np.unique(result)) == 2
+
+    result = graph.cut_threshold(labels, g, thresh)
     assert len(np.unique(result)) == 1