Fix bug in persim.bottleneck calculation (#75)

* Add issue test

* Remove modificatin of ds

* Bump version, fix broken badges in docs

* Bump version, add tests

* Modify test for matching shape

* Remove old travis badge

RELEASE.txt

@@ -1,3 +1,6 @@
+0.3.4 
+    - Fix bug of Issue #70 (https://github.com/scikit-tda/persim/issues/70).
+
 0.3.3
     - Fix plotting methods of Persistence Landscapes, add doc strings.
     - Update to notebooks.

docs/index.rst

@@ -1,4 +1,4 @@
-|PyPI version| |Downloads| |Build Status| |Codecov| |License: MIT|
+|PyPI version| |Downloads| |Codecov| |License: MIT|
 
 Persim is a Python package for many tools used in analyzing Persistence Diagrams.  It currently includes implementations of most of the popular methods of working with persistence diagrams, including
 
@@ -50,12 +50,10 @@ Documentation
     notebooks/Persistence Landscapes and Machine Learning
 
 
-.. |Downloads| image:: https://pypip.in/download/persim/badge.svg
+.. |Downloads| image:: https://img.shields.io/pypi/dm/persim 
     :target: https://pypi.python.org/pypi/persim/
 .. |PyPI version| image:: https://badge.fury.io/py/persim.svg
    :target: https://badge.fury.io/py/persim
-.. |Build Status| image:: https://travis-ci.org/scikit-tda/persim.svg?branch=master
-   :target: https://travis-ci.org/scikit-tda/persim
 .. |Codecov| image:: https://codecov.io/gh/scikit-tda/persim/branch/master/graph/badge.svg
    :target: https://codecov.io/gh/scikit-tda/persim
 .. |License: MIT| image:: https://img.shields.io/badge/License-MIT-yellow.svg

persim/_version.py

@@ -1 +1 @@
-__version__ = "0.3.3"
+__version__ = "0.3.4"

persim/bottleneck.py

@@ -27,9 +27,9 @@ def bottleneck(dgm1, dgm2, matching=False):
 
     Parameters
     -----------
-    dgm1: Mx(>=2) 
+    dgm1: Mx(>=2)
         array of birth/death pairs for PD 1
-    dgm2: Nx(>=2) 
+    dgm2: Nx(>=2)
         array of birth/death paris for PD 2
     matching: bool, default False
         if True, return matching infromation and cross-similarity matrix
@@ -47,15 +47,13 @@ def bottleneck(dgm1, dgm2, matching=False):
     """
 
     return_matching = matching
-
     S = np.array(dgm1)
     M = min(S.shape[0], S.size)
     if S.size > 0:
         S = S[np.isfinite(S[:, 1]), :]
         if S.shape[0] < M:
             warnings.warn(
-                "dgm1 has points with non-finite death times;"+
-                "ignoring those points"
+                "dgm1 has points with non-finite death times;" + "ignoring those points"
             )
             M = S.shape[0]
     T = np.array(dgm2)
@@ -64,8 +62,7 @@ def bottleneck(dgm1, dgm2, matching=False):
         T = T[np.isfinite(T[:, 1]), :]
         if T.shape[0] < N:
             warnings.warn(
-                "dgm2 has points with non-finite death times;"+
-                "ignoring those points"
+                "dgm2 has points with non-finite death times;" + "ignoring those points"
             )
             N = T.shape[0]
 
@@ -101,7 +98,7 @@ def bottleneck(dgm1, dgm2, matching=False):
 
     # Step 2: Perform a binary search + Hopcroft Karp to find the
     # bottleneck distance
-    ds = np.sort(np.unique(D.flatten()))[0:-1] # Everything but np.inf
+    ds = np.sort(np.unique(D.flatten()))  # [0:-1]  # Everything but np.inf
     bdist = ds[-1]
     matching = {}
     while len(ds) >= 1:
@@ -118,18 +115,18 @@ def bottleneck(dgm1, dgm2, matching=False):
             matching = res
             ds = ds[0:idx]
         else:
-            ds = ds[idx + 1::]
+            ds = ds[idx + 1 : :]
 
     if return_matching:
         matchidx = []
-        for i in range(M+N):
+        for i in range(M + N):
             j = matching["{}".format(i)]
             d = D[i, j]
             if i < M:
                 if j >= N:
-                    j = -1 # Diagonal match from first persistence diagram
+                    j = -1  # Diagonal match from first persistence diagram
             else:
-                if j >= N: # Diagonal to diagonal, so don't include this
+                if j >= N:  # Diagonal to diagonal, so don't include this
                     continue
                 i = -1
             matchidx.append([i, j, d])

test/test_distances.py

@@ -2,8 +2,7 @@
 import pytest
 import scipy.sparse as sps
 
-from persim import (bottleneck, gromov_hausdorff, heat, sliced_wasserstein,
-                    wasserstein)
+from persim import bottleneck, gromov_hausdorff, heat, sliced_wasserstein, wasserstein
 
 
 class TestBottleneck:
@@ -103,6 +102,14 @@ def test_repeated(self):
         dist = bottleneck(G, H)
         assert dist == 0.5
 
+    def test_one_diagonal(self):
+        # Issue #70: https://github.com/scikit-tda/persim/issues/70
+        dgm1 = np.array([[0, 10]])
+        dgm2 = np.array([[5, 5]])
+        dist, returned_matching = bottleneck(dgm1, dgm2, matching=True)
+        assert dist == 5.0
+        assert returned_matching.shape[1] == 3
+
 
 class TestWasserstein:
     def test_single(self):