What's Changed

• feat: splits are now rectified so they can be applied in LoadRequest.chunks/requests by @selmanozleyen in #222
• chore: remove concat_strategy by @ilan-gold in #226
• chore: deprecate n_iters by @selmanozleyen in #223
• chore: remove deprecated public ChunkSampler by @selmanozleyen in #224
• feat: allow merge argument for DatasetCollection.add_adatas by @ilan-gold in #198
• ci: bump codecov-action to v6 to fix signature verification by @timtreis in #228
• chore: rename LoadRequests.chunks to requests and allow ndarray by @selmanozleyen in #225
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #229
• feat: categorical sampler by @selmanozleyen in #213
• fix: rename CategoricalSampler to ClassSampler by @selmanozleyen in #232
• feat: DatasetCollection.obs by @selmanozleyen in #230
• feat: support in-memory data by @ilan-gold in #231
• fix: handle different input data types by @ilan-gold in #233

New Contributors

• @timtreis made their first contribution in #228

Full Changelog: v0.1.6...v0.2.0

What's Changed

• fix: ci warnings by @ilan-gold in #210
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #208
• perf: internal integer indexing by @selmanozleyen in #209
• fix: pre-release job CI by @ilan-gold in #217
• chore: update speed graphic by @ilan-gold in #215
• fix: correct url by @ilan-gold in #218
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #211
• docs: Documentation redesign by @Zethson in #216
• chore: changlelog by @ilan-gold in #219

Full Changelog: v0.1.5...v0.1.6

What's Changed

• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #205
• chore: eoss badge by @ilan-gold in #206
• fix: indexing indices and data separately because their chunk grids may differ by @ilan-gold in #207

Full Changelog: v0.1.4...v0.1.5

What's Changed

• Feat: groupby per dataset by @selmanozleyen in #186
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #197
• fix: sampler docs by @ilan-gold in #125
• fix: headers by @ilan-gold in #203
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #202
• fix: centralize zarr version check by @ilan-gold in #204
• perf: write directly to an output buffer by @ilan-gold in #200
• chore: add preshuffling docs by @ilan-gold in #199

Full Changelog: v0.1.3...v0.1.4

What's Changed

• feat: deprecate ChunkSampler in favor of RandomSampler and SequentialSampler while introducing with replacement sampling by @selmanozleyen in #143
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #183
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #185
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #187
• chore: set up pre-testing, clean up hatch by @ilan-gold in #190
• chore: citation by @ilan-gold in #191
• chore: prepare for release 0.1.3 by @ilan-gold in #192

Full Changelog: v0.1.2...v0.1.3

What's Changed

• fix: remove TypeAliasType from nitpick_ignore by @ilan-gold in #174
• chore: add Python 3.14 support by @selmanozleyen in #175
• fix: sampler configuration docs by @ilan-gold in #179
• fix: torch<2.11 compat by @ilan-gold in #180
• fix: cleaner compat for torch<=2.11 by @ilan-gold in #181

Full Changelog: v0.1.1...v0.1.2

What's Changed

• Fix progressbar descriptions by @felix0097 in #166
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #167
• fix: torch 2.11 compat by @ilan-gold in #168
• fix: exclude torch 2.11 that is broken with cupy by @ilan-gold in #170

Full Changelog: v0.1.0...v0.1.1

What's Changed

• chore: clarify warning by @ilan-gold in #149
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #152
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #153
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #154
• feat: io rng handling by @selmanozleyen in #147
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #158
• feat: Add distributed version of ChunkSampler by @felix0097 in #150
• chore: typos, formatting and refactor by @selmanozleyen in #145
• chore: paramtrize main test by var subset by @ilan-gold in #157
• Feat: Clean up params for zarr writing by @felix0097 in #156
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #162
• Feat: Supply size string for n_obs_per_dataset by @felix0097 in #159
• perf: use pinned memory with preload_to_gpu by @ilan-gold in #161
• Update README by @felix0097 in #165
• chore: prepare for 0.1.0 release by @ilan-gold in #164

New Contributors

• @pre-commit-ci[bot] made their first contribution in #152

Full Changelog: v0.0.8...v0.1.0

What's Changed

• Fix: ChunkSampler yields empty splits when remainder is less than batch_size and drop_last by @selmanozleyen in #132
• fix: no more np.random.default_rng().shuffle(batch_indices) by @selmanozleyen in #137
• chore: changelog for rng by @ilan-gold in #139
• fix: yq version by @ilan-gold in #141
• fix: go back to latest yq by @ilan-gold in #142
• Change [tool.pytest.ini_option] to [tool.pytest.ini_options] in pyproject.toml by @selmanozleyen in #134
• feat: Add var handling + validation by @felix0097 in #133
• Bug fix: Loader.__len__ returns wrong length by @felix0097 in #144
• fix: rng handling with workers by @selmanozleyen in #135
• chore: update by @ilan-gold in #148

Full Changelog: v0.0.7...v0.0.8

What's Changed

• fix: support 3d data by @ilan-gold in #126
• perf: add concat_strategy argument to Loader by @ilan-gold in #124
• feat: Make it optional to provide splits for Sampler in LoadRequest by @felix0097 in #127
• Clarify docs on chunk ordering and refactor _get_relative_obs_indices by @selmanozleyen in #128
• perf: use downcasting for sparse matrices when writing to disk by @ilan-gold in #129
• fix: small docs fixes by @ilan-gold in #131

Full Changelog: v0.0.6...v0.0.7