Quarter offset support for cftime #2702

jwenfai · 2019-01-24T12:12:15Z

Closes Implement quarterly frequencies for cftime #2663
Tests added
Fully documented, including whats-new.rst for all changes and api.rst for new API

Waiting for reviews before editing whats-new.rst.

more bugs fixed, cleaned.

…, cleaned.

Resample v2 clean

PEP8

@max-sixty

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Not cleaned.

Get PEP8 changes from Ouranosinc.

…-v2-clean # Conflicts: # xarray/tests/test_cftimeindex_resample.py

@max-sixty

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Cleaned.

# Conflicts: # xarray/core/resample.py # xarray/tests/test_cftimeindex_resample.py

Resample v2 clean

Updating to latest xarray.

Update to latest xarray.

…a() in groupby.py. Pandas cannot drop indices made up of CFTime objects, so integer indices were swapped in for dropping then swapped back out once NAs are dropped.

…after more tests. 5578 out of 5920 tests passed. Pre-cleaning.

@max-sixty

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Not cleaned.

@max-sixty

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Cleaned.

…a() in groupby.py. Pandas cannot drop indices made up of CFTime objects, so integer indices were swapped in for dropping then swapped back out once NAs are dropped.

…after more tests. 5578 out of 5920 tests passed. Pre-cleaning.

…after more tests. 5578 out of 5920 tests passed. Cleaned.

…_get_range_edges.

…n common.py. Cleaned up print statements. Modified tests that were written under the assumption that CFTimeIndex cannot be resampled so that the tests now pass.

…ex.py. Removed unused code from resample_cftime.py. Removed tests that raise error when resampling CFTimeIndex. Removed temp files.

…range is off is the month of start date % 3 != 0

…PEP8 and other coding style conventions.

…sets and resampling quarter offsets. Default values added to CFTimeGrouper.

…opefully.

…o quarter-offset

jwenfai · 2019-01-27T20:50:58Z

I rebased, seems like the failing checks are the same as the ones affecting the resample pull request.

spencerkclark · 2019-01-27T21:14:41Z

Sorry @jwenfai, things still don't look quite right (we should only see commits related to the addition of quarterly offsets here). If it's easier, feel free to start fresh from a new PR.

seems like the failing checks are the same as the ones affecting the resample pull request.

As @shoyer mentioned in #2593, the tests should be fixed if you merge the latest master branch, now that #2720 has been merged.

spencerkclark · 2019-01-27T21:36:47Z

xarray/tests/test_cftimeindex_resample.py

+
+@pytest.mark.parametrize('freq', [
+    '2M', '3M',
+    '2Q-JUN', '3Q-JUL',


Ah, I see you've made some changes related to resampling with quarterly frequencies here. My apologies, I didn't notice that before. That's great, but it would be nice to hold off on that until #2593 has been officially merged. It looks like most of what you're looking to add in this PR is independent of resampling (you only need to update the definition of is_super_daily and how the default arguments are defined to support resampling with quarterly frequencies, right?).

Most of the changes are indeed not resample related, except for is_super_daily in get_time_bins and similar code in CFTimeGrouper.

I'll start a new pull request that doesn't touch any of the resample code.

Sounds great, thanks -- that will make things easier to review 👍. Sorry for my confusion here.

jwenfai and others added 30 commits November 9, 2018 16:58

First implementation of resampling for CFTimeIndex.

daa3a71

First implementation of resampling for CFTimeIndex, cleaned.

f9f3347

First implementation of resampling for CFTimeIndex, cleaned.

0950505

First implementation of resampling for CFTimeIndex, cleaned.

89f418a

First implementation of resampling for CFTimeIndex.

39c9d11

First implementation of resampling for CFTimeIndex,

073b8e0

more bugs fixed, cleaned.

First implementation of resampling for CFTimeIndex, test file written.

193c4c4

First implementation of resampling for CFTimeIndex, test file written…

2c97738

…, cleaned.

First implementation of resampling for CFTimeIndex, test file written…

9993ed9

…, cleaned.

First implementation of resampling for CFTimeIndex, test file written…

f01745c

…, cleaned.

First implementation of resampling for CFTimeIndex, test file written…

ffbf265

…, cleaned.

Merge pull request #1 from jwenfai/resample-v2-clean

e64fedb

Resample v2 clean

Docstrings for resample_cftime.py written. Upsample still not fixed.

2850dd5

Fixed PEP8 and test parametrization.

770b778

PEP8

181e82c

Merge pull request #3 from Ouranosinc/PEP8

5a41ee2

PEP8

Test file fixes and other optimizations (2018-12-16 @spencerclark and…

6b948c5

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Not cleaned.

Merge pull request #1 from Ouranosinc/master

97c0948

Get PEP8 changes from Ouranosinc.

Merge remote-tracking branch 'origin/resample-v2-clean' into resample…

63d25ab

…-v2-clean # Conflicts: # xarray/tests/test_cftimeindex_resample.py

Test file fixes and other optimizations (2018-12-16 @spencerclark and…

05af869

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Cleaned.

Merge branch 'resample-v2-upsample' into resample-v2-clean

85f1a84

# Conflicts: # xarray/core/resample.py # xarray/tests/test_cftimeindex_resample.py

Merge pull request #4 from jwenfai/resample-v2-clean

2e8ced3

Resample v2 clean

Merge branch 'master' into master

4317c69

Merge pull request #2 from Ouranosinc/master

e7deeb2

Updating to latest xarray.

Merge pull request #3 from Ouranosinc/master

f9ac1a1

Update to latest xarray.

_get_range_edges logic changed to emulate latest version of pandas.

8505ca9

Simplified resampling logic (errors persist). Pre-cleaning.

a495c2d

Simplified resampling logic (error persists). Cleaned.

ad65ef0

Simplified resampling logic (error persists). Fixed first_items.dropn…

5775e11

…a() in groupby.py. Pandas cannot drop indices made up of CFTime objects, so integer indices were swapped in for dropping then swapped back out once NAs are dropped.

Simplified resampling logic (error persists). Logic slightly altered …

e1902fe

…after more tests. 5578 out of 5920 tests passed. Pre-cleaning.

jwenfai and others added 24 commits January 27, 2019 13:45

Fixed PEP8 and test parametrization.

c293201

PEP8

bc46146

Test file fixes and other optimizations (2018-12-16 @spencerclark and…

5f64e75

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Not cleaned.

Test file fixes and other optimizations (2018-12-16 @spencerclark and…

54e0b1e

… 2018-12-05 @max-sixty GitHub reviews for resample-v2-clean pull request). Cleaned.

_get_range_edges logic changed to emulate latest version of pandas.

a359be6

Simplified resampling logic (errors persist). Pre-cleaning.

f37eaac

Simplified resampling logic (error persists). Cleaned.

ffdbc84

Simplified resampling logic (error persists). Fixed first_items.dropn…

c42e9cc

…a() in groupby.py. Pandas cannot drop indices made up of CFTime objects, so integer indices were swapped in for dropping then swapped back out once NAs are dropped.

Simplified resampling logic (error persists). Logic slightly altered …

a8d2ed8

…after more tests. 5578 out of 5920 tests passed. Pre-cleaning.

Simplified resampling logic (error persists). Logic slightly altered …

a09e4c7

…after more tests. 5578 out of 5920 tests passed. Cleaned.

Precise cftime arithmetic. Reduced overall test time. Added test for …

59c09d8

…_get_range_edges.

Added default values for closed and label args of resample function i…

5890dff

…n common.py. Cleaned up print statements. Modified tests that were written under the assumption that CFTimeIndex cannot be resampled so that the tests now pass.

Added back replace['dayofwk'] = -1 to cftime_offsets.py and cftimeind…

57bb50c

…ex.py. Removed unused code from resample_cftime.py. Removed tests that raise error when resampling CFTimeIndex. Removed temp files.

Optimizations as per pydata#2593

25de49f

Simple test for non-standard calendars added and documentation updated.

e061d34

Simple test for non-standard calendars added and documentation updated.

bdd39d8

Quarter offset implementation. Incorrect as first month of generated …

ba8b0d6

…range is off is the month of start date % 3 != 0

Added loffset support to CFTimeIndex resampling. Better adherence to …

a5ccc8c

…PEP8 and other coding style conventions.

Added loffset support to CFTimeIndex resampling. Better adherence to …

92226c1

…PEP8 and other coding style conventions.

Support datetime.timedelta objects for loffset. Improved test coverage.

fcb6581

Quarter offsets now supported for cftime. Added tests for quarter off…

0eab794

…sets and resampling quarter offsets. Default values added to CFTimeGrouper.

Quarter offsets now supported for cftime. Added tests for quarter off…

9c70506

…sets and resampling quarter offsets. Default values added to CFTimeGrouper.

Rebased quarter-offset branch to pydata/master. Conflicts resolved, h…

3f4d8c5

…opefully.

Merge remote-tracking branch 'ouranosinc-upstream/quarter-offset' int…

d21ad03

…o quarter-offset

Merge branch 'pydata-master' into quarter-offset

823d88b

spencerkclark reviewed Jan 27, 2019

View reviewed changes

jwenfai closed this Jan 27, 2019

jwenfai deleted the quarter-offset branch January 27, 2019 22:05

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Quarter offset support for cftime #2702

Quarter offset support for cftime #2702

Uh oh!

jwenfai commented Jan 24, 2019

Uh oh!

jwenfai commented Jan 27, 2019

Uh oh!

spencerkclark commented Jan 27, 2019

Uh oh!

spencerkclark Jan 27, 2019 •

edited

Loading

Uh oh!

jwenfai Jan 27, 2019

Uh oh!

spencerkclark Jan 27, 2019

Uh oh!

Uh oh!

Uh oh!

Quarter offset support for cftime #2702

Quarter offset support for cftime #2702

Uh oh!

Conversation

jwenfai commented Jan 24, 2019

Uh oh!

jwenfai commented Jan 27, 2019

Uh oh!

spencerkclark commented Jan 27, 2019

Uh oh!

spencerkclark Jan 27, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jwenfai Jan 27, 2019

Choose a reason for hiding this comment

Uh oh!

spencerkclark Jan 27, 2019

Choose a reason for hiding this comment

Uh oh!

Uh oh!

spencerkclark Jan 27, 2019 •

edited

Loading