Add default zero constructor for QuadPrecision#84
Conversation
Most python numeric types have default constructors, and pandas relies on it when building string representations of data frames
|
Fixes #83 |
|
Looks like you need to declare Pandas as a test dependency. |
|
Good point. I think I added that in the right place now. |
|
It looks like there are still failures on the tests for older CPUs and more obscure architectures. Sorry this is so tricky! I think you might need to use importorskip to skip the test if pandas isn't importable. |
|
btw, you can enable actions on your fork so you don't need to wait on me to hit the approve button over here |
|
It turns out the test dependencies weren't being installed in some of the tests. I fixed that and now it's passing on my repo except for the build docs job which looks like it's missing an access token. |
SwayamInSync
left a comment
SwayamInSync
left a comment
There was a problem hiding this comment.
Just added some small nits and test to cover longdouble backend.
Thanks @ejeffrey LGTM!!
|
CI is all green, merging this 🚀 |
3 participants
Most python numeric types have default constructors, and pandas relies on it when building string representations of data frames