pydata / xarray / 3658
96%

DEFAULT BRANCH: master
Ran
Jobs 6
Files 71
Run time 11min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
3658

push

travis-ci

jhamman 
Groupby-like API for resampling (#1272)

* Implement basic functionality by adding "DataArrayResample" and "DatasetResample" subclasses of their equivalent "GroupBy" implementation

* Re-factor old resample logic to separate method

* Adding test cases for new api

* Adding more DataArray resample tests

* Adding test_dataset test cases

* Update to use proxy __resample_dim__ on resampling ops

* Re-factor proxy resampling dimension and catch rare error if used as actual resampling dimension name

* BUG: Fixed loss of attrs on DatasetResample even when keep_attrs=True

* Update docs to add information about new resampling api

* Adding 'Whats new' entry

* Tweak auxiliary groupby apply and reduce methods

* Squash bugs from py.test following rebase to master

Un-do _combine temporary debugging output

* Fixing typo in groupby.py

* Add a test for resampling which includes timezones in the datetime stamps

* Rolling back the timezone tests... this will need to be tackled separately

* Re-factored resample into it's own module for easier maintenance

* Adding support for 'count' via old api

* Re-organizing new vs old resampling api tests

* Expanded old-vs-new api tests for Dataset to replace deprecated tests

* Consolidated old reasmpling api tests into new-vs-old for dataarray

* Wrapping old api test invocations with pytest.warns

* Added stub tests for upsampling

* Update documentation with upsampling stub

* Factor out a Resample object and add initial up-sampling methods - bfill, pad, asfreq

* Add interpolation up-sampling to DataArray

* Refine DataArray upsampling interpolation and extend to Dataset

* Fix wrong time dimension length on test cases for upsampling

* First initial revisions to @shoyer's comments; before modifying implementation of DataArrayResample

* Tweaks to resample.py to lean on super-methods

* Implementing interpolation test cases

* BUG: Fix asfreq only returning 1D data in nd case

* Add pad/asfreq upsampling tests

* Add a check if old/new api is mixed; remove old api details from resample docstring

* Fix an old bug in datetime components of timeseries doc

* Tweaking time-series doc

* Added what's new entry

* Drop existing non-dimension coordinates along resample dimension

* Update seaborn to v0.8 to fix issues with default plot styles

* nearest-neighbor up-sampling now relies on re-index instead of interpolate

* Adding nearest upsampling test; tweaked inference of re-indexing dimensions to avoid always having to compute means

* Move what's new entry to breaking changes

* Updating docs in breaking changes with example and link to timeseries page

* BUG: Fixing creating merged coordinates for Dataset upsampling case

* Remove old notice about resampling api updates

* Applying shoyer's clean-up of figuring out valid coords on Dataset up-sampling

* Add note about monotonicity assumption before interpolation

* fix some pep8 and comments

* More informative error message when resampling/interpolating dask arrays

* Fix flake8

* Fixing issues with test cases, including adding one for upsampling dask arrays

* Clean up scipy imports

* Adding additional tweaks to cover scipy/numpy dependencies in tests

17277 of 17726 relevant lines covered (97.47%)

5.28 hits per line

Jobs
ID Job ID Ran Files Coverage
1 3658.1 (CONDA_ENV=py27-min) 71
74.69
 Travis Job 3658.1
2 3658.2 (CONDA_ENV=py27-cdat+pynio) 71
96.82
 Travis Job 3658.2
3 3658.3 (CONDA_ENV=py34) 71
74.65
 Travis Job 3658.3
4 3658.4 (CONDA_ENV=py35) 71
94.17
 Travis Job 3658.4
5 3658.5 (CONDA_ENV=py36) 71
94.39
 Travis Job 3658.5
10 3658.10 (CONDA_ENV=py36-bottleneck-dev) 71
93.2
 Travis Job 3658.10
Source Files on build 3658
Coverage ∆ File Lines Relevant Covered Missed Hits/Line