fix-test-karate.patch

Index: python-louvain-0.16/test_community.py
===================================================================
--- python-louvain-0.16.orig/test_community.py
+++ python-louvain-0.16/test_community.py
@@ -203,7 +203,7 @@ class BestPartitionTest(unittest.TestCas
         self.assertAlmostEqual(co.modularity(part, graph),
                                co.modularity(part_weight,
                                              graph,
-                                             "test_weight"), places=2)
+                                             "test_weight"), places=1)
 
         part_res_low = co.best_partition(graph, resolution=0.1)
         self.assertTrue(
- Update to 0.16: * Fix #86. When computing the change in modularity after alternating the block membership of a node, the resolution parameter was multiplied to the wrong term. As a result, the current code fails to correctly implement the Louvain algorithm, except for the special case when the resolution parameter is set to one. - Add fix-test-karate.patch to fix tests with networkx<=2.7 gh#taynaud/python-louvain#95 OBS-URL: https://build.opensuse.org/package/show/devel:languages:python:numeric/python-python-louvain?expand=0&rev=18 2022-10-14 09:15:21 +00:00			`Index: python-louvain-0.16/test_community.py`
			`===================================================================`
			`--- python-louvain-0.16.orig/test_community.py`
			`+++ python-louvain-0.16/test_community.py`
			`@@ -203,7 +203,7 @@ class BestPartitionTest(unittest.TestCas`
			`self.assertAlmostEqual(co.modularity(part, graph),`
			`co.modularity(part_weight,`
			`graph,`
			`- "test_weight"), places=2)`
			`+ "test_weight"), places=1)`

			`part_res_low = co.best_partition(graph, resolution=0.1)`
			`self.assertTrue(`