Once this is fixed, how do we get the new version on conda forge so that travis will pull it correctly for hera_cal? Or should we move to pip installing this repo from github?
Coverage increased (+0.02%) to 47.387% when pulling 38cb1a241f7515b0321d866a18724aff4b3f5d07 on fix-BeamAlm-set-params into 09327f5d348e33b252a4ae49f0f106cfa31d5d87 on master.
Use truncation division which was intended and broken by making aipy both python 2 and 3 compatible.
This is currently breaking hera_cal: https://travis-ci.org/HERA-Team/hera_cal/builds/417473551?utm_source=github_status&utm_medium=notification