0.0.3

2024 release

This release includes significant updates over version 0.0.2a0, including:

• New support for Fused Unbalanced Gromov-Wasserstein (FUGW) methods
• New support for Individualized Neural Tuning (INT) methods
• New support for Shared Response Model (SRM) methods
• New support for Hyperalignment methods
• New accompanying examples and documentation
• Accompanying bug fixes and dependency updates

What's Changed

• Changing OT backend from POT to OTT by @thomasbazeille in #48
• [ENH] Bump to nilearn v0.9 by @emdupre in #51
• Patch fix on #51 by @emdupre in #52
• Fixed tests by @bthirion in #57
• [Bugfix] Remove deprecated normalize option in RidgeCV by @pbarbarant in #59
• Trying new configuration by @bthirion in #58
• Updates for newer ott-jax interface by @emdupre in #53
• [MAINT] Packaging by @emdupre in #60
• [MAINT] Fix tests by @emdupre in #61
• [MAINT] Use Nilearn's hierarchical kmeans implentation by @emdupre in #63
• [INFRA] Remove Circle in favor of GHA by @emdupre in #64
• [FMT] Re-introduce flake8, solve existing fmt errors by @emdupre in #65
• [INFRA] First pass at doc builder GHA by @emdupre in #66
• [INFRA] Update sphinx conf for newer version, correct warnings by @emdupre in #69
• [INFRA] Untrack nilearn, set 10.2 as minimum version by @emdupre in #76
• [ENH] Add fastSRM functionality by @emdupre in #70
• [INFRA] nilearn updates by @emdupre in #79
• Add Parcel and Searchlight Hyperalignment methods by @denisfouchard in #74
• [FEAT] Add fugw interface by @pbarbarant in #83
• [INFRA] Try doc deployment from jupyter-book by @emdupre in #90
• Create a deploy action by @pbarbarant in #85
• [DOC] Restrict deploy job to run only on the main branch by @pbarbarant in #97
• [DOC] Initial commit of furo theme by @emdupre in #99
• [INFRA] Expanding supported Python versions by @emdupre in #101

New Contributors

• @pbarbarant made their first contribution in #59
• @denisfouchard made their first contribution in #74

Full Changelog: 0.0.2a0...0.0.3

0.0.2a0

Release accompanying benchmark paper

This release accompanied the paper "An empirical evaluation of functional alignment using inter-subject decoding" (doi: 10.1016/j.neuroimage.2021.118683).

It includes several significant improvements over the previous release, including

• Support for Recursive Neighbor Agglomeration (ReNA)
• Improved usage patterns, including clearer errors and metrics
• Better library testing

What's Changed

• Unlock ReNA if nilearn version is enough by @thomasbazeille in #35
• Reput hierarchical_kmeans in fmralign by @thomasbazeille in #39
• [DOC] Note recommended region size for OT convergence by @emdupre in #41
• [doc] improve error when calling unrecognized alignment method by @emdupre in #42
• [fix] reduce warnings by @emdupre in #43
• fix: update get_data to get_fdata method by @emdupre in #44
• [ENH] Adding score metrics by @emdupre in #40
• [FIX, ENH, MRG] Improve clustering functions, tests, by @thomasbazeille in #46
• Version2 by @thomasbazeille in #47

Full Changelog: 0.0.1a0...0.0.2a0

0.0.1a0

Initial release

Initial release of the fmralign library 🎉

What's Changed

• Installation updates by @thomasbazeille in #2
• Alignment methods and tests by @thomasbazeille in #3
• [MRG] Implemented diagonal estimator to solve Issue #7 by @thomasbazeille in #10
• [MRG] Add circle ci by @thomasbazeille in #11
• [MRG] Pairwise alignment module, tests and examples by @thomasbazeille in #4
• [MRG] Pass parallelization back-end as an argument by @thomasbazeille in #21
• Make POT an optional dependency by @thomasbazeille in #23
• Fixing deprecated argument in CircleCI env creation by @thomasbazeille in #24
• [MRG] pandas upgrade by @bthirion in #26
• Fixing deprecated argument in Circle CI configuration by @thomasbazeille in #27
• [MRG] Template alignment module and tests by @thomasbazeille in #5
• [MRG] Parcellation(#6) and solve issues #16 and #17 by @thomasbazeille in #29
• [MRG] Sphinx, narrative documentation... by @thomasbazeille in #30
• [MRG] Alignment methods benchmarking example and documentation by @thomasbazeille in #31
• [MRG] API Improvements to solve #22 and #19 by @thomasbazeille in #32
• Readme Release Update by @thomasbazeille in #33

Full Changelog: https://github.com/Parietal-INRIA/fmralign/commits/0.0.1a0