| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2026-03-10 | 5.6 kB | |
| v0.7.5 source code.tar.gz | 2026-03-10 | 7.9 MB | |
| v0.7.5 source code.zip | 2026-03-10 | 9.1 MB | |
| Totals: 3 Items | 17.0 MB | 0 | |
What's Changed 🚀
💥 Breaking Changes
- perf!: lazy-load lance to speed up
import daft@desmondcheongzx (#6328) - feat!: add ddof parameter to stddev (default ddof=1) @aaron-ang (#6142)
✨ Features
- feat(io): support numeric range expansion in glob patterns @singularityDLW (#6127)
- feat(dashboard): add query result preview tab @samstokes (#6327)
- feat: Add
read_textAPI to support reading text files @plotor (#6111) - feat: Add protocol aliases for IOConfig @universalmind303 (#6252)
- feat(observability): Improve distributed operator names @srilman (#6324)
- feat(metrics): add node_plan_id for distributed stats attribution @cckellogg (#6332)
- feat: adds support for daft native extensions via stable C ABI @rchowell (#6301)
- feat(lance): refine distributed scalar index support with new types and parameters @huleilei (#5817)
- feat!: add ddof parameter to stddev (default ddof=1) @aaron-ang (#6142)
- feat: use human-readable query names instead of UUIDs @samstokes (#6310)
- feat(metrics): add node phase attribute for multi-phase pipeline nodes @cckellogg (#6290)
- feat(dashboard): add wall-clock and CPU duration to operator cards @samstokes (#6300)
- feat: Add Python Function data source info in explain @plotor (#6119)
- feat(dashboard): physical plan tree with live operator stats @samstokes (#6299)
- feat(io): implement write_sql with SQLDataSink and explicit dtype support @huleilei (#5979)
- feat(dashboard): add interactive tree visualizer for query plans @samstokes (#6295)
🐛 Bug Fixes
- fix(video): add missing pillow dependency for video keyframes @desmondcheongzx (#6355)
- fix(dashboard): request notification permission before enabling notifications @samstokes (#6362)
- fix(dashboard): make dashboard subscriber non blocking @cckellogg (#6336)
- fix(dashboard): ensure query_end fires when generator is abandoned early @samstokes (#6326)
- fix(tests): make Ray conda UDF tests robust to unavailable Python micro versions @cckellogg (#6341)
- fix(sql): reject duplicate CTE names in WITH clause @singularityDLW (#6331)
- fix: Use Tokio FS for Flight Server Reads @srilman (#6308)
- fix(io): add connection pool semaphore to Azure Blob Storage backend @singularityDLW (#6305)
- fix(dashboard): sort extra stats alphabetically to prevent flicker @samstokes (#6309)
- fix: Support Union and Optional in
DataType.infer_from_type@srilman (#6304) - fix: make
max_concurrencycontrol coroutine concurrency for async UDFs @colin-ho (#6302) - fix: Fix native executor closing @colin-ho (#6282)
- fix(docs): Remove duplicate Voice AI analytics listing in SUMMARY.md @copilot-swe-agent[bot] (#6274)
🚀 Performance
- perf: Drop tasks early in distributed limits @srilman (#6350)
- perf(image): make image ops run in parallel via rayon @universalmind303 (#6333)
- perf(upload): use a zero copy approach for url_upload @universalmind303 (#6334)
- perf(parquet): late materialization via RowFilter and RowSelection @desmondcheongzx (#6311)
- perf!: lazy-load lance to speed up
import daft@desmondcheongzx (#6328)
♻️ Refactor
- refactor: decouple extension ABI from arrow-rs with Arrow C Data Interfaces @rchowell (#6337)
- refactor(schema): change Field.name from String to Arc\<str> @desmondcheongzx (#6351)
- refactor(deps): remove 11 dependencies by inlining builder dependencies @universalmind303 (#6364)
- refactor(arrow2): goodbye arrow2 @universalmind303 (#6363)
- refactor(arrow2): migrate remaining parquet code to arrow-rs @universalmind303 (#6361)
- refactor(arrow2): remove parquet2 from daft @desmondcheongzx (#6339)
- refactor(arrow2): remove parquet2 read pipeline, make arrow-rs the default reader @desmondcheongzx (#6335)
- refactor(arrow2): add arrow-rs parquet reader @desmondcheongzx (#6294)
- refactor(arrow2): daft-csv crate @universalmind303 (#6298)
- refactor(arrow2): daft-functions-serde @universalmind303 (#6303)
- refactor(arrow2): introduce DaftParquetMetadata adapter to decouple crates from parquet2 @desmondcheongzx (#6292)
- refactor(arrow2): refactor daft-json to arrow-rs @universalmind303 (#6297)
- refactor(arrow2): move the auto conversion logic into array init @universalmind303 (#6289)
📖 Documentation
- docs(optimization): add join strategies guide @desmondcheongzx (#6367)
- docs: add custom catalog implementation guide @desmondcheongzx (#6366)
- docs(skills): add Claude Code compatible guides and update README @Jay-ju (#6116)
🔧 Maintenance
- chore: update Slack invite links across docs and README @everettVT (#6377)
- chore: Replace RepartitionSpec::IntoPartitions with custom path @srilman (#6349)
- chore: Improve Common Crawl Benchmark @srilman (#6307)
- chore(tests): migrate internal usages of daft.udf to cls/func @universalmind303 (#6348)
- chore(dashboard): add README with build and architecture notes @samstokes (#6323)
- chore(lints): add workspace lints inheritance to missing subcrates @yew1eb (#6330)
- chore: Fix some typo issues in the code or docs @plotor (#6320)
- chore(deps): bump the all group with 4 updates @dependabot[bot] (#6314)
- chore: refactor pipeline node ownership in start @srilman (#6259)
- chore: Move EXPLAIN ANALYZE to Python @srilman (#6254)
⬆️ Dependencies
- chore(deps): bump the all group with 4 updates @dependabot[bot] (#6314)
Full Changelog: https://github.com/Eventual-Inc/Daft/compare/v0.7.4...v0.7.5