0.3.1

What's Changed

• Fix typo in filter method docstings by @mnrozhkov in #250
• Skip if not SQLite Improvements by @dtulga in #254
• Autodetect Studio branch by @dreadatour in #253
• Autodetect Studio branch fix for 'main' branch by @dreadatour in #257
• Removing metastore argument from Client.parse_url() by @ilongin in #256
• Autodetect Studio branch fix for 'main' branch by @dreadatour in #258
• Parallel UDF optimizations by @dreadatour in #211

Full Changelog: 0.3.0...0.3.1

0.3.0

What's Changed

• wds - misc changes by @dmpetrov in #232
• Merge 'parent' and 'name' columns into 'path' by @rlamy in #101

Full Changelog: 0.2.18...0.3.0

0.2.18

What's Changed

• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #238
• Adding DataChain.column(...) and fixing functions and types by @ilongin in #226

Full Changelog: 0.2.17...0.2.18

0.2.17

What's Changed

• Update readme by @dmpetrov in #233
• Arrow nrows fix by @dberenbaum in #221
• only combine final step for limit by @mattseddon in #230
• Fix renaming object or normal signal with .mutate() by @ilongin in #217
• Fixing too many files open, and adding reconnect by @dtulga in #229

Full Changelog: 0.2.16...0.2.17

0.2.16

What's Changed

• improve efficiency of examples by @mattseddon in #214
• fix select then distinct chain by @mattseddon in #213
• rename DataChain's create_empty to from_records by @mattseddon in #215
• do not modify datachain max limit in show by @mattseddon in #225
• Rename cleanup_temp_tables to cleanup_tables in warehouse and catalog by @amritghimire in #218

Full Changelog: 0.2.15...0.2.16

0.2.15

What's Changed

• Arrow improvements by @dberenbaum in #126
• prevent cryptic error messages when running llm claude examples by @mattseddon in #194
• remove reference to missing notebook by @mattseddon in #193
• Fix for nested lists of models in schema by @ilongin in #195
• Support for DataChain.batch_map() by @dberenbaum in #191
• renamed clip.py by @dberenbaum in #201
• validation error handling improved by @volkfox in #203
• modified Mistral prompt and changed to the DataModel by @volkfox in #205
• make datachain show respect existing limit by @mattseddon in #206
• JSON tutorial by @volkfox in #207
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci in #197
• Move 'create_pre_udf_table' function to warehouse module by @dreadatour in #187

New Contributors

• @pre-commit-ci made their first contribution in #197

Full Changelog: 0.2.14...0.2.15

0.2.14

What's Changed

• Fixing test warnings by @dtulga in #158
• Update DataChain.subtract() to work without legacy file signals by @rlamy in #157
• Optimizations: low-hanging fruits by @dreadatour in #178
• from_values with array of arrays by @dmpetrov in #183
• remove collect_one from example by @mattseddon in #186
• fixing a missed import for codegen schemas by @volkfox in #184
• avoid instantiating filesystem for path operations by @skshetry in #176
• Remove get_possibly_stale_jobs from metastore by @amritghimire in #189

New Contributors

• @amritghimire made their first contribution in #189

Full Changelog: 0.2.13...0.2.14

0.2.13

What's Changed

• DataChain.from_storage: add last_modified and is_latest to the columns by @skshetry in #165
• fix for using new column from .mutate() in .order_by() by @ilongin in #171
• Renaming File.write() to File.save() by @ilongin in #172
• storage: index as a dir if no glob by @shcheklein in #108

Docs

• first shot at LLM eval tutorial by @volkfox in #145
• Update README.rst by @volkfox in #161
• docs: update README.rst by @eltociear in #168

Maintenance

• skip mypy hook on pre-commit.ci by @skshetry in #164
• ci: disable azure and gs remote tests on macOS by @skshetry in #174
• ci: run s3 tests on Windows, be more careful while skipping by @skshetry in #175
• fix test for ch wrt datetime precision by @skshetry in #169
• Adding tests for exporting image files and File.write() by @ilongin in #149

New Contributors

• @eltociear made their first contribution in #168

Full Changelog: 0.2.12...0.2.13

0.2.12

What's Changed

• Python API to manage the dataset registry by @dreadatour in #29
• cli: hide subcommands from the listing by @skshetry in #79
• datachain: rename include_sys kwarg to sys by @skshetry in #69
• Adding DataChain.export_files(...) by @ilongin in #30
• Update cv tutorial: fashion_product_images by @mnrozhkov in #62
• Add and clean up docstrings in datachain api by @dberenbaum in #63
• docs: fix invalid python code inside docstrings by @skshetry in #85
• Hide traceback for xfails in Studio test runs by @rlamy in #87
• Rename UDF to UDFStep for clarity, and remove from root namespace by @rlamy in #88
• Fix mutate() by @dmpetrov in #78
• update pytest-servers to 0.5.5 by @mattseddon in #94
• Remove vendored-code-specific folders by @dtulga in #95
• Rename repository references to datachain by @dtulga in #93
• do not overwrite version with None in DatasetQuery constructor by @mattseddon in #92
• always include sys signals by @skshetry in #81
• Add more UniqueId fields by @rlamy in #90
• Added more generalize SignalsSchema.;get_signals() method instead of get_file_signals(...) by @ilongin in #86
• Added input params to distinct() by @ilongin in #96
• Fix for order_by with sub signals by @ilongin in #82
• Remove legacy signals in from_storage() by @rlamy in #72
• Updates to examples by @dberenbaum in #77
• More docs updates by @dberenbaum in #100
• Add 'update' param to DataChain.from_storage method by @dreadatour in #99
• Fix repository reference in Notebook by @dtulga in #105
• fix(ux): remove reference to DatasetQuery by @shcheklein in #104
• datachain: implement to_parquet by @skshetry in #97
• File refactor by @dberenbaum in #102
• fixing regressions from switching to ModelStore.add() by @volkfox in #109
• add ModelStore to top level imports by @dmpetrov in #112
• add truncate option to show and update default width of output by @mattseddon in #116
• merge/join: exclude sys signals by @skshetry in #120
• Added descending parameter to DataChain.order_by(...) by @ilongin in #122
• remove get_value() from DataModel by @dmpetrov in #119
• Add file modes for binary/text by @dberenbaum in #107
• remove docstring from DataModel.pydantic__init_subclass by @skshetry in #123
• Examples cleanup by @dberenbaum in #111
• rename ModelStore.add() to register() by @dmpetrov in #113
• datachain: generalize data access functions into collect(), and collect_flatten by @skshetry in #121
• Add nrows for partial parsing of csv/parquet by @dberenbaum in #124
• Update index.md by @volkfox in #128
• Picture for getting started by @volkfox in #127
• moving pic to the right place by @volkfox in #131
• cleanup signal refs in examples by @dberenbaum in #129
• cleanup api reference index by @dberenbaum in #130
• Fix for text and images files export by @ilongin in #135
• update computer vision quick start example by @mattseddon in #136
• update computer vision image example by @mattseddon in #139
• Huggingface test updates and bug fix by @dberenbaum in #140
• Readme update by @dmpetrov in #133
• readme: fix link to image by @dmpetrov in #143
• Update badge by @skshetry in #144
• don't depend on datachain from PATH to exec processes by @skshetry in #118
• dc: try to fix dataset_stats for DataChain.from_storage() generated dataset by @skshetry in #151

New Contributors

• @dreadatour made their first contribution in #29
• @mnrozhkov made their first contribution in #62

Full Changelog: 0.2.11...0.2.12

0.2.11

What's Changed

• cleanup model store/registry by @dberenbaum in #74
• slice nested signals by @dberenbaum in #75
• To pandas - hierarchical multi header by @dmpetrov in #22
• Use cloudpickle for parallel UDF processing by @dtulga in #65

Full Changelog: 0.2.10...0.2.11