Circular imports #30
Circular imports #30
Conversation
There was a problem hiding this comment.
The more I think about it, the more I become convinced that it's best to simply delete the arclines dependency and put the algorithms in the PYPIT package. This would have several advantages:
- We wouldn't have to replicate all of the robust fitting code now in utils.py
- We could use some of the other PYPIT functionality (like msgs)
- Given that PYPIT is being modularised, it should become possible to simply "from pypit import arclines".
I can't think of any negatives.... In any case, now seems like the most appropriate time to discuss this and make a decision. We started talking about this in Issue #26, so let's discuss this as an Issue.
| @@ -6,6 +6,8 @@ | |||
| import numpy as np | |||
| import pdb | |||
|
|
|||
| from arclines import utils as al_utils | |||
There was a problem hiding this comment.
I'll just make a note about this here... I tend to agree with Kyle in this post about restricting the use of as to only when it's absolutely necessary (e.g. different naming of packages/utilities in python 2 vs 3).
|
merging for now, but arclines is headed into PYPIT proper |
Remove all calls to the pypit package
to avoid circular imports. This meant
duplicating a few of the methods in
pypit's arutils (e.g robust_polyfit).
It will be wise in the future to begin tagging
releases of arclines.