Closed beringueb closed 8 months ago
Looks good but we should refer to the correct version of syslibrary. Did you push this before the modifications in syslibrary were merged to the master? Or having them to the master is not enough and there should be a new syslibrary version released?
I don't think syslibrary has versions ? But yes the reason why the tests have failed is because it looks for the wrong version
yeah, I see in the repository that there is no released version... but I have this memory that syslibrary had to be released in order for mflike to be pip installable... is that right @xgarrido ?
syslibrary
should push a new version to pypi and we should increase the requirement here
I'm still trying to understand the eval
function but I guess we should also change the default values of theses parameters https://github.com/simonsobs/syslibrary/blob/master/syslibrary/syslib_mflike.py#L44 in syslibrary
. Maybe we can also require no default values...
I think the default are correct now "t" instead of "tt". I will check about pushing syslib to pypi.
Ok I was still looking to your previous PR and you've added this one https://github.com/simonsobs/syslibrary/pull/18
We need syslibrary
to be on pypi but it's already the case https://pypi.org/project/syslibrary/ So someone @beringueb , @mgerbino need to push a version 0.1.1
at least to pypi
Martina wil take care of it tomorrow morning. I'll then bump the requirements in setup.py and we should be good to merge
Changing the keys to cal_pars to reflect the fact that they are a_lm (or map) level calibraation. This is now more consistent and makes with what has been changed to syslibrary here: https://github.com/simonsobs/syslibrary/pull/17