Resolving Issue 85: TypeError in coatingthermal with old version of numpy
Resolving the issue mentioned in Issue 85 regarding the behavior of np.flip in numpy versions before 1.15. Modern np.flip has a default axis option while older versions required you to give an axis specifically. We are only flipping 1-dimensional arrays so I have replaced all np.flip with np.flipud which is equivalent to calling np.flip(m, 0). This should solve the issue. However, locally I could not confirm this as I was having segmentation fault error. The local CI test also couldn't pass due to a problem with all package. So hopefully, the reviewers can test this branch in the appropriate place. I don't see a reason for why this would fail with older numpy versions now.
Merge request reports
Activity
added 1 commit
- 33522b91 - Removing comments that were left over from previous version.
added 1 commit
- 019ad778 - Using explicit np.flipud instead of providing axis=0 or relying on default behavior
mentioned in issue #86 (closed)
added 12 commits
-
019ad778...a101f0c4 - 10 commits from branch
gwinc:master - 0e3a1775 - Using explicit np.flipud instead of providing axis=0 or relying on default behavior
- a02feb9e - Merge remote-tracking branch 'origin/master'
-
019ad778...a101f0c4 - 10 commits from branch
mentioned in commit 87859500