New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
MAINT: simplify discrete calls to get_robustcov_results #5234
Conversation
Codecov Report
@@ Coverage Diff @@
## master #5234 +/- ##
==========================================
- Coverage 86.07% 84.09% -1.99%
==========================================
Files 698 613 -85
Lines 118309 97619 -20690
Branches 13131 10657 -2474
==========================================
- Hits 101840 82093 -19747
+ Misses 13707 13118 -589
+ Partials 2762 2408 -354
Continue to review full report at Codecov.
|
use_transparams=use_transparams, | ||
disp=False) | ||
|
||
assert result.use_t is use_t |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
All four of the use_t = True
, cov_type='nonrobust'
cases fail under master. @josef-pkt is that intentional?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@josef-pkt insights here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@bashtage any idea if the existing behavior is correct? If so, what am I missing?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@josef-pkt when you get a chance this is a specifically-your-attention-needing thing. Is the existing behavior intentional?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This might be a bug as in #5331
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This might be a bug as in #5331
Neat. Before I dig into this, is the simplification+unification in this PR something you're open to?
The general idea looks good. There are some tricky parts to the internal temporary transformation, but I don't remember at the moment what the status is. e.g. #3747 |
e4a87ae
to
047fa14
Compare
047fa14
to
78edac6
Compare
78edac6
to
63fba0c
Compare
I'm pretty sure this is either incorrect or fragile because of the extra dispersion parameter |
and I guess it computes the sandwich cov_params twice, the first time incorrectly |
63fba0c
to
47192a3
Compare
47192a3
to
8dda7c7
Compare
get_robustcov_results
is called inLikelihoodModelResults.__init__
. Avoid re-calling it to make things less-stateful.