Review dependencies on interpolations package

[mmcky]

The interpolations package tries to maintain bleeding edge integration with numba and currently supports numba>=0.49. This causes issues with the lecture series dependence on anaconda releases as numba trails in anaconda.

We should see if we can add jitted interpolation functions to QuantEcon.py to replace dependence on interpolations

[albop]

I've got no objection here, but if it's a problem it's easy to add backward compatibility to interpolation.py. The breaking change just comes from numba moving the location of it's core types.

[mmcky]

hey @albop -- if you're willing to add in support for numba that is contained in the latest anaconda distribution that would be great. 👍 -- Currently that is numba=0.48.

@najuzilu is working to add some simpler interpolation functions to QuantEcon.py so we will likely add them (+ a testing framework) as a backup.

[albop]

Yes, it's no big deal. I've spent a lot of time recently doing conda packaged like you taught me. Will check back on a few hours

On Mon, May 11, 2020, 9:49 AM mmcky [email protected] wrote:

hey @albop https://github.com/albop -- if you're willing to add in support for numba that is contained in the latest anaconda distribution that would be great. 👍 -- Currently that is numba=0.48.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/QuantEcon/lecture-python/issues/198#issuecomment-626534032, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACDSKJUNRXJR7EY25XALG3RQ6U2NANCNFSM4M3ZDOPQ .