Change Log¶
v1.4.0 (2019-07-05)¶
Closed issues:
- Why features are iterated in a for loop? #243
Merged pull requests:
- Two-sided outlier filtering mode #249 (gbordyugov)
- Create .zappr.yaml #248 (perploug)
- Correct documentation #245 (shansfolder)
v1.3.8 (2018-09-10)¶
Merged pull requests:
- Changed chi-square test, removed frequencies computation #240 (daryadedik)
v1.3.7 (2018-09-04)¶
Merged pull requests:
- Added observed and expected frequencies to chi-square statistics #239 (daryadedik)
- Stop warning about NaNs #238 (aaron-mcdaid-zalando)
- stop overriding the warning level #237 (aaron-mcdaid-zalando)
v1.3.6 (2018-08-06)¶
Merged pull requests:
- If the number of *finite* samples is too small, then the data isn’t valid for analysis #236 (aaron-mcdaid-zalando)
v1.3.5 (2018-08-02)¶
Merged pull requests:
- Re-wrote chi-square, removed dropping buckets. #234 (daryadedik)
- check if there are no p values to correct #233 (gbordyugov)
v1.3.3 (2018-07-26)¶
Merged pull requests:
- Chi-squared test for the variant split check #228 (daryadedik)
v1.3.2 (2018-07-23)¶
Merged pull requests:
- Fix minor typos #231 (sdia)
- custom deepcopy() method for ‘StatisticalTest’, to save some memory #230 (aaron-mcdaid-zalando)
- Updated old multiple correction documentation #229 (daryadedik)
- contributing.rst: Improve the release procedure to ensure that the up… #226 (aaron-mcdaid-zalando)
v1.2.6 (2018-07-01)¶
Merged pull requests:
- Ensure that outlier detection works if there is NaN in the data #225 (aaron-mcdaid-zalando)
- More powerful derived kpis #222 (aaron-mcdaid-zalando)
v1.2.5 (2018-06-22)¶
Merged pull requests:
- Counting bugfix and save memory #224 (aaron-mcdaid-zalando)
- Fix for the possibility that both variances are zero #221 (aaron-mcdaid-zalando)
v1.2.4 (2018-05-31)¶
Merged pull requests:
- Remove null analysis results from the analysis results files #219 (daryadedik)
v1.2.3 (2018-05-30)¶
Merged pull requests:
- Removed deep copy of the data in statistical test construction #218 (daryadedik)
v1.2.1 (2018-05-29)¶
Merged pull requests:
- Added merge_with class method for merging two multiple test suite results and tests #216 (daryadedik)
- List of filtered columns as filtered_columns metadata information #215 (daryadedik)
v1.1.0 (2018-05-24)¶
Merged pull requests:
- Experiment data restructure #213 (daryadedik)
- Original corrected results #212 (daryadedik)
v1.0.1 (2018-04-23)¶
Merged pull requests:
- Fixed docstring #211 (daryadedik)
- raise ValueError on zero pooled std for power calculations #210 (gbordyugov)
- Changed structure for statistics without correction #209 (daryadedik)
v1.0.0 (2018-03-22)¶
Merged pull requests:
- Finish Documentation #204 (shansfolder)
- Fix logging sga error logging #203 (igusher)
- Project Headache #194 (shansfolder)
v0.6.13 (2018-03-15)¶
Implemented enhancements:
- Applying bins to data frames #165
Fixed bugs:
Merged pull requests:
- Wrap sga in try catch #202 (igusher)
- Multiple correction method module #201 (shansfolder)
- Adapted util module and util unit tests #199 (daryadedik)
- Adapt early stopping #198 (daryadedik)
- Adapt statistics.py #197 (shansfolder)
- Adapt experiment module #196 (shansfolder)
- Make result classes JSON serializable #195 (shansfolder)
- Results data structure #193 (shansfolder)
- fixed small typos in percentiles and doc text #191 (daryadedik)
- fixing sample size estimation #188 (gbordyugov)
v0.6.12 (2018-01-24)¶
Merged pull requests:
- Doc update #186 (shansfolder)
- AXO-103 include variance in delta / group-sequential reports #185 (gbordyugov)
v0.6.11 (2018-01-23)¶
Merged pull requests:
- Axo-91 bug fix sga #184 (shansfolder)
- added code coverage badge and reformatted README.rst a bit #183 (mkolarek)
v0.6.7 (2018-01-10)¶
Closed issues:
- Group Sequential - Percentile Issue #176
Merged pull requests:
- Increase version to 0.6.7 #181 (shansfolder)
- fixed last command in “Deploying to PyPI” part of contributing.rst #180 (mkolarek)
- Extended multiple correction for group sequential, added doc for multiple correction. #179 (daryadedik)
- Fix information fraction calculation #177 (shansfolder)
v0.6.6 (2017-11-27)¶
Closed issues:
- Infinitely large confidence intervals produced by group_sequential_delta() #172
Merged pull requests:
- Merging dev to master for new release #175 (mkolarek)
- AXO-35 implemented estimate_sample_size() for estimating sample size … #174 (mkolarek)
- Fix two-sided alpha value in power analysis #173 (shansfolder)
- Docs/update contrib doc #171 (mkolarek)
- Add some parameter checks #170 (shansfolder)
- Make applying bins to data frames more agreeable #169 (gbordyugov)
- OCTO-2181: Implement over time analysis. Time-based SGA #164 (daryadedik)
v0.6.3 (2017-10-24)¶
Merged pull requests:
- OCTO-2214 Bugfix: Capping information fraction #163 (shansfolder)
- OCTO-2088: Implement multiple testing correction in ExpAn #161 (daryadedik)
- OCTO-1044 Improve readthedoc #160 (shansfolder)
- OCTO-1933 Subgroup analysis #159 (shansfolder)
- release 0.6.2 #156 (mkolarek)
- OCTO-1920, OCTO-1968, OCTO-1969 Refactor binning #155 (shansfolder)
v0.6.2 (2017-08-29)¶
Fixed bugs:
- Result statistics in Baeysian methods #142
Closed issues:
Merged pull requests:
- make development requirements open ended #154 (mkolarek)
- Octo 1930 implement quantile filtering #153 (mkolarek)
- Not use empty list for method parameter #152 (shansfolder)
- OCTO-1971 Add variational inference for early stopping #150 (shansfolder)
- Updated intro documentation covering delta methods. #149 (daryadedik)
- Release v0.6.1 #148 (shansfolder)
- Merge pull request #137 from zalando/dev #147 (shansfolder)
- Add static html file from intro doc for v0.6.1 #146 (shansfolder)
v0.6.1 (2017-08-08)¶
Implemented enhancements:
- Optimizing the control flow from
ExperimenttoResults#82 - more meaningful dict keys for results #139 (gbordyugov)
Fixed bugs:
- reenable means and bounds functions on Results object #9
Closed issues:
- Results.to_json() implementation not flexible #65
- Results.to_json() doesn’t support trend() results #64
Merged pull requests:
- Documentation updates for Expan 0.6.x. Covers OCTO-1961, OCTO-1970 #145 (daryadedik)
- Fix delta/alpha model para inconsistency #144 (shansfolder)
- Small improvement on default type of report_kpi_names #140 (shansfolder)
- slightly different json structure for results #138 (gbordyugov)
- merging dev to master #137 (gbordyugov)
v0.6.0 (2017-07-26)¶
Closed issues:
Merged pull requests:
- Octo 1616 no experimentdata #134 (gbordyugov)
- Attempt to fix pickling bug #133 (shansfolder)
- Stan models compilation, exceptions catch, unit tests adaptation. #131 (daryadedik)
- Added try-finally block for the compulsory clean-up of .pkl compiled models #130 (daryadedik)
- OCTO-1837 fixed to_json() #129 (gbordyugov)
v0.5.3 (2017-06-26)¶
Implemented enhancements:
- Weighted KPIs is only implemented in regular delta #114
Fixed bugs:
- Assumption of nan when computing weighted KPIs #119
- Weighted KPIs is only implemented in regular delta #114
- Percentiles value is lost during computing group_sequential_delta #108
Closed issues:
- Failing early stopping unit tests #85
Merged pull requests:
- Release new version 0.5.3 #127 (mkolarek)
- OCTO-1804: Optimize the loading of .stan model in expan. #126 (daryadedik)
- Test travis python version #125 (shansfolder)
- OCTO-1619 Cleanup ExpAn code #124 (shansfolder)
- OCTO-1748: Make number of iterations as a method argument in _bayes_sampling #123 (daryadedik)
- OCTO-1615 Use Python builtin logging instead of our own debugging.py #122 (shansfolder)
- OCTO-1711 Support weighted KPIs in early stopping #121 (shansfolder)
- Fixed a few bugs #120 (shansfolder)
- OCTO-1614 cleanup module structure #115 (shansfolder)
- OCTO-1677 : fix missing .stan files #113 (gbordyugov)
- Bump version 0.5.1 -> 0.5.2 #112 (mkolarek)
v0.5.2 (2017-05-11)¶
Implemented enhancements:
- OCTO-1502: cleanup of call chains #110 (gbordyugov)
Merged pull requests:
- OCTO-1502 support **kwargs for four delta functions #111 (shansfolder)
- new version 0.5.1 #107 (mkolarek)
v0.5.1 (2017-04-20)¶
Implemented enhancements:
- Derived KPIs are passed to Experiment.fixed_horizon_delta() but never used in there #96
Merged pull requests:
- updated CONTRIBUTING.rst with deployment flow #106 (mkolarek)
- OCTO-1501: bugfix in Results.to_json() #105 (gbordyugov)
- OCTO-1502 removed variant_subset parameter… #104 (gbordyugov)
- OCTO-1540 cleanup handling of derived kpis #102 (shansfolder)
- OCTO-1540: cleanup of derived kpi handling in Experiment.delta() and … #97 (gbordyugov)
- Small refactoring #95 (shansfolder)
- Merge dev to master for v0.5.0 #94 (mkolarek)
v0.5.0 (2017-04-05)¶
Implemented enhancements:
Fixed bugs:
- Experiment.sga() assumes features and KPIs are merged in self.metrics #87
- pctile can be undefined in
Results.to\_json\(\)#78
Closed issues:
- Results.to_json() => TypeError: Object of type ‘UserWarning’ is not JSON serializable #77
- Rethink Results structure #66
Merged pull requests:
- new dataframe tree traverser in to_json() #92 (gbordyugov)
- updated requirements.txt to have ‘greater than’ dependencies instead … #89 (mkolarek)
- pip version requirement #88 (gbordyugov)
- Test #86 (s4826)
- merging in categorical binning #84 (gbordyugov)
- Add documentation of the weighting logic #83 (jbao)
- Early stopping #80 (jbao)
- a couple of minor cleanups #79 (gbordyugov)
- Merge to_json() changes #75 (mkolarek)
- Feature/early stopping #73 (jbao)
v0.4.5 (2017-02-10)¶
Fixed bugs:
- Numbers cannot appear in variable names for derived metrics #58
Merged pull requests:
- Feature/results and to json refactor #74 (mkolarek)
- Merge to_json() and prob_uplift_over_zero changes #72 (mkolarek)
- regex fix, see https://github.com/zalando/expan/issues/58 #70 (gbordyugov)
v0.4.4 (2017-02-09)¶
Implemented enhancements:
- Add argument assume_normal and treatment_cost to calculate_prob_uplift_over_zero() and prob_uplift_over_zero_single_metric() #26
- host intro slides (from the ipython notebook) somewhere for public viewing #10
Closed issues:
- migrate issues from github enterprise #20
Merged pull requests:
v0.4.2 (2016-12-08)¶
Fixed bugs:
- frequency table in the chi square test doesn’t respect the order of categories #56
Merged pull requests:
v0.4.1 (2016-10-18)¶
Merged pull requests:
- small doc cleanup #55 (jbao)
- Add comments to cli.py #54 (igusher)
- Feature/octo 545 add consolidate documentation #53 (mkolarek)
- added os.path.join instead of manual string concatenations with ‘/’ #52 (mkolarek)
- Feature/octo 958 outlier filtering #50 (mkolarek)
- Sort KPIs in reverse order before matching them in the formula #49 (jbao)
v0.4.0 (2016-08-19)¶
Closed issues:
- Support ‘overall ratio’ metrics (e.g. conversion rate/return rate) as opposed to per-entity ratios #44
Merged pull requests:
v0.2.5 (2016-05-30)¶
Implemented enhancements:
- Implement __version__ #14
Closed issues:
- upload full documentation! #1
Merged pull requests:
- implement expan.__version__ #19 (pangeran-bottor)
- Mainly documentation changes, as well as travis config updates #17 (robertmuil)
- Update README.rst #16 (pangeran-bottor)
- added cli module #11 (mkolarek)
- new travis config specifying that only master and dev should be built #4 (mkolarek)
v0.2.4 (2016-05-16)¶
Closed issues:
- No module named experiment and test_data #13
Merged pull requests:
v0.2.0 (2016-05-06)¶
Merged pull requests:
- Added detailed documentation with data formats #3 (robertmuil)
* This Change Log was automatically generated bygithub_changelog_generator