Skip to content

Latest commit

 

History

History
385 lines (191 loc) · 17.4 KB

CHANGELOG.md

File metadata and controls

385 lines (191 loc) · 17.4 KB
Raw

CHANGELOG

v0.1.3 (2023-11-25)

Build

  • build: Update .readthedocs.yml (5b2311a)

  • build: add the rtd configuration (cc34ffd)

v0.1.2 (2023-11-25)

Build

  • build: switch back to napoleon (fda2e78)

  • build: add epytext module for doc generation (4f7a8d5)

  • build: update poetry for documents (1e9b2db)

Chore

  • chore: fix line endings on changelog (f4b5643)

Fix

  • fix: handle series data in datasets (353f50c)

  • fix: Not all functions of doenut.py were being exported. (304047f)

Unknown

  • complete basic docs (e24bc40)

  • doc: yet more docstrings (d0f1a4c)

  • doc: yet more docstrings (8d29f05)

  • doc: more docstrings (6c53e8c)

  • doc: formatting (9d01fbb)

  • doc: many more docstrings (f902b94)

  • doc: wip on tidying docstrings for publishing (dafff2f)

  • doc: add basic generated docs (ffdfbc4)

  • add default code of conduct / contrib files (#31) (5242f44)

  • 26 proper logging (#30)

  • feat: added basic logging framework

  • feat: more logging implemented

  • feat: finish adding basic logging across package. (2eb036d)

v0.1.1 (2023-11-18)

Build

Unknown

  • 22 check all the docstrings help docs (#28)

  • fix: column_selector should allow response only filtering

  • fix: scaling helper function has wrong default for optional

  • doc: A lot of docstrings

  • doc: better readme

  • chore: add semantic release (74a32ff)

v0.1.0 (2023-11-17)

Build

  • build: prepare for v0.1.0 release (77049c1)

Unknown

  • fix designer (267b3ff)

  • add tune_model to averaged_model. Oh and black now supports ipynb (b825e26)

  • Set scaling to not break POLS by only scaling inputs by default (46b3ef2)

  • 18 refactor into classes (#19)

  • move plotting into its own file

  • Doenut is no longer calulated.

  • add Model and AveragedModel. Make the function formerly known as calulate use the new model code

  • make scaling per-column

  • add documentation to model code

  • restore distinction between where scaling can be done.

  • Add first pass at calulating a selective model + a test for it.

  • Don't need to pass the response key in, we know it.

  • tidy up some imports

  • hive model set functionality from averaged model off into its own class

  • rework of model sets

  • got input_selector working on averaged model

  • further work on classification. Sort out sequencing of R2/Q2 calculations

  • added autotune test

  • autotune now calls averagemodel directly

  • add filterable dataframe helper

  • rename and move terms into new package

  • Tweaked data so averaged duplicate is actually different to original data. implemented duplicate removal and averaging in filtered data frame

  • add average / remove duplicates to DataFrameSet

  • make more sense of the dataset object. duplicates handled OK, need to finish averaging code

  • add builder pattern to set functions in DataSet, tidy up typing.

  • had an annoying idea to make this better.

  • start of insane rewrite of classes

  • Basic dataset and filtered dataset implemented and tests added

  • rework (again) to match builder pattern

  • add scaling modifier

  • rename / sort the modifiers into their own module, add duplicate handling modifiers

  • Add tests for duplicate handling

  • work on updating model classes to use new dataset

  • further work on fixing models

  • two tests left

  • one test remaining

  • added a bunch of documentation to the code.

  • make all the modifiers return a new object.

  • split data_set into data_set and modifiable_data_set

  • allow for deciding if responses should be scaled

  • autotune now works

  • clean up defunct code from doenut.py (mainly)

  • test on python 3.10

  • hopefully fix test on python 3.10

  • further typing fixes for earlier pythons

  • missed some

  • further tidying up. Updated the Test 2 example

  • update manual (53f21c9)

  • 4 remove doepy (#12)

  • implement new version of full_fact

  • revert full fact to taking a dictionary

  • add test for fractional model

  • add wrapper around fract fact (b258ffa)

  • enable pytest in github actions (#11) (ec95d49)

  • Merge pull request #9 from ellagale/6-create-basic-workflow-demonstration-regression-test

6 create basic workflow demonstration regression test (b83f58b)

  • work on docstrings, reformating (d17d374)

  • reformat + lint tests (3a7f59f)

  • add last of model validation (4e95ba0)

  • complete second set of tests (a868b7d)

  • further work on doenut test. Numbers need validation. (479a043)

  • work on removing warnings (1ca0b4c)

  • remove missing parameter errors (ebcd86c)

  • test basic parsimonious model (ef20d0e)

  • test basic full model (763c1c0)

  • test basic R2/Q2 calculation (a0b559b)

  • update a couple of functions for new pandas changes (fac2b0c)

  • add initial test (16101ae)

  • switch to poetry from setuptools (29ca60a)

  • remove pycharm files from repo (d032cea)

  • Merge pull request #8 from ellagale/3-switch-license-to-mit

switch license to MIT (c50314c)

  • switch license to MIT (b5fedc8)

  • add rework warning to readme (f52a3ad)

  • Merge pull request #2 from cwoac/main

Turn doenut into a module pt1 (71680da)

Create LICENSE (f174e8a)

Added in error plots (15fd46d)