Skip to content
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

Update imports, use nogil version of sqrt #18557

Merged
merged 5 commits into from
Dec 4, 2017

Conversation

jbrockmendel
Copy link
Member

No description provided.

@codecov
Copy link

codecov bot commented Nov 29, 2017

Codecov Report

Merging #18557 into master will decrease coverage by 0.01%.
The diff coverage is 88.88%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master   #18557      +/-   ##
==========================================
- Coverage   91.35%   91.33%   -0.02%     
==========================================
  Files         164      164              
  Lines       49802    49800       -2     
==========================================
- Hits        45496    45485      -11     
- Misses       4306     4315       +9
Flag Coverage Δ
#multiple 89.13% <88.88%> (-0.01%) ⬇️
#single 40.81% <55.55%> (-0.07%) ⬇️
Impacted Files Coverage Δ
pandas/core/tools/datetimes.py 84.43% <ø> (-0.06%) ⬇️
pandas/core/indexes/datetimes.py 95.71% <100%> (-0.01%) ⬇️
pandas/tseries/offsets.py 96.9% <50%> (ø) ⬆️
pandas/io/gbq.py 25% <0%> (-58.34%) ⬇️
pandas/core/frame.py 97.81% <0%> (-0.1%) ⬇️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 32f562d...703a54b. Read the comment docs.

@codecov
Copy link

codecov bot commented Nov 29, 2017

Codecov Report

Merging #18557 into master will increase coverage by <.01%.
The diff coverage is 88.88%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master   #18557      +/-   ##
==========================================
+ Coverage   91.44%   91.45%   +<.01%     
==========================================
  Files         157      157              
  Lines       51449    51447       -2     
==========================================
+ Hits        47048    47049       +1     
+ Misses       4401     4398       -3
Flag Coverage Δ
#multiple 89.32% <88.88%> (+0.02%) ⬆️
#single 40.6% <55.55%> (-0.11%) ⬇️
Impacted Files Coverage Δ
pandas/core/tools/datetimes.py 84.43% <ø> (-0.06%) ⬇️
pandas/core/indexes/datetimes.py 95.68% <100%> (-0.01%) ⬇️
pandas/tseries/offsets.py 96.9% <50%> (ø) ⬆️
pandas/io/gbq.py 25% <0%> (-58.34%) ⬇️
pandas/core/frame.py 97.81% <0%> (-0.1%) ⬇️
pandas/plotting/_converter.py 65.25% <0%> (+1.81%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bd9a3e0...e743c1e. Read the comment docs.

@jreback
Copy link
Contributor

jreback commented Nov 29, 2017

would it not be better to expose this as an definition in a single .pxd, maybe call it cmath.pxd (along with other ones), then simply import centrally from there (or could also just add to util.pxd).

@jreback jreback added the Clean label Nov 29, 2017
@jbrockmendel
Copy link
Member Author

would it not be better to expose this as an definition in a single .pxd, maybe call it cmath.pxd (along with other ones), then simply import centrally from there (or could also just add to util.pxd).

I see the upside to this but also like transparent/direct imports. Not sure there is enough content to justify a cmath.pxd; what else would go in there?

I'll open an issue with cython to see if they can fix the declaration on their end too and make this (eventually) a non-issue.

For purposes of expediency, would it be easier if I revert the sqrt bits now and get the easy other fixups closed down?

@jreback
Copy link
Contributor

jreback commented Nov 29, 2017

i would put it in util.pxd

the issue is it’s very likely that future changes will accidentally just use the wrong import and just do it with d

@jreback jreback closed this Nov 29, 2017
@jreback jreback reopened this Nov 29, 2017
@@ -32,11 +32,10 @@ cdef double NaN = <double> np.NaN
cdef inline int int_max(int a, int b): return a if a >= b else b
cdef inline int int_min(int a, int b): return a if a <= b else b

from util cimport numeric
from util cimport numeric, sqrt

cdef extern from "../src/headers/math.h":
int signbit(double) nogil
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

prob worthwhile adding any of the imports from src/headers/math here as well (prob not very many)

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

prob worthwhile adding any of the imports from src/headers/math here as well (prob not very many)

Agreed; its been added to the src todo list. Need to move on for now.

@jbrockmendel
Copy link
Member Author

conda HTTPError 502 on Appveyor. Will wait until late-evening so re-push as the build queue is usually short then.

@jreback
Copy link
Contributor

jreback commented Nov 29, 2017

looks fine. prefer that you update all of these in this pr.

@jbrockmendel
Copy link
Member Author

Travis error in 3.6 TestDataFramePlots.test_parallel_coordinates_with_sorted_labels

@jbrockmendel
Copy link
Member Author

prefer that you update all of these in this pr.

I opened an issue on the cython tracker and based on the response I think we should stick a pin in that until we have a better idea what's going on.

@jreback
Copy link
Contributor

jreback commented Nov 30, 2017

are these in cython 0.26 or greater?

@jbrockmendel
Copy link
Member Author

are these in cython 0.26 or greater?

All my local builds should be on 0.27.1

@jreback
Copy link
Contributor

jreback commented Dec 3, 2017

and what is the perf affect of this
specifically related to the revert of the nogil sqrt changes

@jbrockmendel
Copy link
Member Author

specifically related to the revert of the nogil sqrt changes

The "revert" is reverting from an earlier version of the PR, i.e. keeps the status quo in master unchanged. This should be nothing but import cleanup.

@jreback jreback added this to the 0.22.0 milestone Dec 4, 2017
@jreback jreback merged commit e99cb9c into pandas-dev:master Dec 4, 2017
@jreback
Copy link
Contributor

jreback commented Dec 4, 2017

thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants