Move warnings import to the beginning of knn.py file

Use isfinite instead of isinf and isnan
Use catch_warnings context manager instead of "always" with simplefilter

1 files changed, 10 insertions, 24 deletions

diff --git a/src/python/gudhi/point_cloud/knn.py b/src/python/gudhi/point_cloud/knn.py
index 0724ce94..de5844f9 100644
--- a/src/python/gudhi/point_cloud/knn.py
+++ b/src/python/gudhi/point_cloud/knn.py
@@ -8,6 +8,7 @@
 #   - YYYY/MM Author: Description of the modification
 
 import numpy
+import warnings
 
 # TODO: https://github.com/facebookresearch/faiss
 
@@ -257,14 +258,9 @@ class KNearestNeighbors:
             if ef is not None:
                 self.graph.set_ef(ef)
             neighbors, distances = self.graph.knn_query(X, k, num_threads=self.params["num_threads"])
-            if numpy.any(numpy.isnan(distances)):
-                import warnings
-                warnings.simplefilter("always")
-                warnings.warn("NaN value encountered while computing 'distances'", RuntimeWarning)
-            if numpy.any(numpy.isinf(distances)):
-                import warnings
-                warnings.simplefilter("always")
-                warnings.warn("Overflow value encountered while computing 'distances'", RuntimeWarning)
+            with warnings.catch_warnings():
+                if not(numpy.all(numpy.isfinite(distances))):
+                    warnings.warn("Overflow/infinite value encountered while computing 'distances'", RuntimeWarning)
             # The k nearest neighbors are always sorted. I couldn't find it in the doc, but the code calls searchKnn,
             # which returns a priority_queue, and then fills the return array backwards with top/pop on the queue.
             if self.return_index:
@@ -298,14 +294,9 @@ class KNearestNeighbors:
             if self.return_index:
                 if self.return_distance:
                     distances, neighbors = mat.Kmin_argKmin(k, dim=1)
-                    if torch.isnan(distances).any():
-                        import warnings
-                        warnings.simplefilter("always")
-                        warnings.warn("NaN value encountered while computing 'distances'", RuntimeWarning)
-                    if torch.isinf(distances).any():
-                        import warnings
-                        warnings.simplefilter("always")
-                        warnings.warn("Overflow encountered while computing 'distances'", RuntimeWarning)
+                    with warnings.catch_warnings():
+                        if not(torch.isfinite(distances).all()):
+                            warnings.warn("Overflow/infinite value encountered while computing 'distances'", RuntimeWarning)
                     if p != numpy.inf:
                         distances = distances ** (1.0 / p)
                     return neighbors, distances
@@ -314,14 +305,9 @@ class KNearestNeighbors:
                     return neighbors
             if self.return_distance:
                 distances = mat.Kmin(k, dim=1)
-                if torch.isnan(distances).any():
-                    import warnings
-                    warnings.simplefilter("always")
-                    warnings.warn("NaN value encountered while computing 'distances'", RuntimeWarning)
-                if torch.isinf(distances).any():
-                    import warnings
-                    warnings.simplefilter("always")
-                    warnings.warn("Overflow encountered while computing 'distances'", RuntimeWarning)
+                with warnings.catch_warnings():
+                    if not(torch.isfinite(distances).all()):
+                        warnings.warn("Overflow/infinite value encountered while computing 'distances'", RuntimeWarning)
                 if p != numpy.inf:
                     distances = distances ** (1.0 / p)
                 return distances