Replace dimensions that are equal but not identical. #158
Conversation
|
Whoops, definitely not ready. Talk about a misread of the code! |
|
I think this version should work. The tests pass, and the yt tests pass as well, but alternate suggestions are appreciated. |
ngoldbaum
force-pushed
the
unsan_dim_replace
branch
from
122f92c
to
3964d82
Compare
|
Sorry for taking a couple days to respond. I looked at this and spotted some issues but wanted to understand the issue and code a little more clearly myself so I decided to rewrite this a little bit to make it clearer for me and not respond until I had time to do that. I added a test that fails on master using the existing old JSON unit registry already used by the tests so hopefully we won't regress on this. I'd appreciate it if you or @Xarthisius or whoever else could verify that the change as I have it here does fix the yt issue in the same way as your original version of this pull request. I didn't have time to run the tests locally so they might fail, if so I'll fix that ASAP and ping you. Unfortunately I need to accompany Shawna to the prosthetist this afternoon so I won't be able to do that for a few hours. Hope you all are having a good week and enjoyed scipy :) |
|
I checked-out |
This closes #157 by examining the old-style dimensions, and if they are equal but not identical, replacing them. This allows them to match the dimensions checked in the unit registry later.