Repository navigation
TST: Fix doctest failures - #10874
Conversation
mhvk
left a comment
There was a problem hiding this comment.
Ah, the risk of the very careful [ci skip] that still ran on readthedocs and passed. So, clearly one shouldn't do that. Thanks for noticing and fixing so quickly!!
|
The RTD job only confirms that it can render into HTML without errors or warnings. It does not run any doctest. Hope this clarifies the matter! |
|
Hmmm! Behavior change in numpy-dev . @mhvk , can you please advise? |
|
Wow, that's interesting/weird... I guess little choice but to |
|
Great, thanks so much! Hopefully it will now all pass... |
|
All green (even the numpy-dev job)! |
TST: Fix doctest failures
|
Weirdly it doesn't cause a traceback on 1.16 on the LTS branch with 32bit 3.6, so I just go ahead and ignore the output as a workaround for that branch. (I suppose I could remove the backport, but this is almost at the bottom of the pile and would be more painful that doing a quick workaround) |
Description
This pull request is to fix doctest failures caused by #10848