Releases: AllenInstitute/ipfx
Releases · AllenInstitute/ipfx
v2.1.2
What's Changed
- tests/data/nwb/Ctgf-T2A-dgCre;Ai14-495723.05.02.01.nwb: convert to Git LFS by @t-b in #570
- Fix on premise tests by @t-b in #581
- Get it working again for newer versions of dependencies by @t-b in #582
- Require pinned pynwb version again by @t-b in #580
- Fix and silence warnings by @t-b in #578
- Update packaging to pyproject.toml by @t-b in #571
- .github/workflows: Use latest stable version of python 3.14 and drop support for python 3.9 by @t-b in #587
- Move release CI from circle CI to github actions by @t-b in #584
- 2.1.0 release by @t-b in #588
- Fix release CI by @t-b in #589
- .github/workflows/deploy_release.yml: Regex anchor ^ is not supported… by @t-b in #591
v2.0.0
What's Changed
- Python 3.9/3.11 support, dropping Python 3.6 support, and moving to GitHub Actions by @sheriferson in #559
- 2.0.0 release by @sheriferson in #563
New Contributors
- @sheriferson made their first contribution in #559
Full Changelog: v1.0.8...v2.0.0
Release 1.0.7
Merge pull request #552 from AllenInstitute/release-1.0.7 Release 1.0.7
Release 1.0.5
Bug fixes:
- Converts nwb_version attribute to string if it is in utf-8 encoded bytes.
Release 1.0.4
Changed:
- selects recording rather than sweep epoch by default
- restructures epoch detection to use recording epoch for stim epoch detection
- Use NaNs instead of truncating
Bug fixes:
- Round the duration calculation in
run_feature_vector_extractionso that
vectors of the same length are produced even when floating point approximations
of times are different.
Release 1.0.3
Changed:
- Fixes memory leak in LRU cache
- Adds new tag and error handling for sweeps with NaN's in stimulus
Release 1.0.2
Changed:
- Add features_state information to the pipeline output json
- Improve performance of loading sweeps from NWB2 files by using LRU cache
- More robust error checking when loading time_series
Release 1.0.1
Updates docker image to decouple from AllenSDK with dedicated ipfx docker build