v0.2.33

Changes

✨ New Features

• [FEAT]: sql case/when @universalmind303 (#2591)
• [FEAT] Add comparison of timestamps with same timezone @Vince7778 (#2604)
• [FEAT] Add support for pyiceberg v0.7 @kevinzwang (#2594)
• [FEAT] Make the end argument for .list.slice() optional @desmondcheongzx (#2593)

🚀 Performance Improvements

• [PERF] Add physical plan optimizer and optimization @Vince7778 (#2557)

👾 Bug Fixes

• [BUG]: remove simsimd dependency @universalmind303 (#2605)
• [BUG] Fix parquet reads when a top-level column's final row spans more than one data page @desmondcheongzx (#2586)
• [BUG]: accept "iterable[pa.Table]" for from_arrow @universalmind303 (#2583)

📖 Documentation

• [CHORE] Fix imports on jupyter notebook examples @kevinzwang (#2600)

🧰 Maintenance

• [CHORE] Fix imports on jupyter notebook examples @kevinzwang (#2600)
• [CHORE]: ignore ".zed" directory @universalmind303 (#2595)

v0.2.32

Changes

✨ New Features

• [FEAT] Fix resource accounting in PyRunner @jaychia (#2567)
• [FEAT] Add .str.count_matches() @Vince7778 (#2580)
• [FEAT]: streaming json @universalmind303 (#2582)
• [FEAT]: embedding.cosine_distance function @universalmind303 (#2526)
• [FEAT] Enable buffered iteration on plans @jaychia (#2566)
• [FEAT] Streaming CSV Reads @colin-ho (#2565)
• [FEAT] Support Reading Iceberg Merge-on-Read Position Deletes @kevinzwang (#2563)
• [FEAT]: daft sql @universalmind303 (#2558)
• [FEAT] Initializes daft-sql and defines the daft.sql(..) function. @rchowell (#2559)
• [FEAT] Add image mode casting @Vince7778 (#2562)
• [FEAT] Add tracing to local execution engine @samster25 (#2556)
• [FEAT] Create file when writing dataframe with no rows @kevinzwang (#2540)
• [FEAT] Add string tokenize expression @Vince7778 (#2503)
• [FEAT]: support optional rowgroups to read_parquet @universalmind303 (#2534)
• [FEAT] Add hashjoin, sort, and hashagg ops to new executor @colin-ho (#2530)

👾 Bug Fixes

• [BUG] Fix .str.length() on Unicode strings @Vince7778 (#2579)
• [BUG] Fix table filters with scalar mask @colin-ho (#2542)
• [BUG] Enable sorting bool columns @colin-ho (#2529)

📖 Documentation

• [FEAT] Add .str.count_matches() @Vince7778 (#2580)
• [DOCS]: small update to "scaling-up" @universalmind303 (#2577)
• [DOCS] quickstart-revision @avriiil (#2124)
• [DOCS] Update Ray Dataset link @kevinzwang (#2564)
• [FEAT] Add image mode casting @Vince7778 (#2562)
• [DOCS] Fix notebook CI @jaychia (#2544)
• [FEAT] Add string tokenize expression @Vince7778 (#2503)
• [DOCS] Do not generate PDFs @jaychia (#2539)
• [DOCS] Build PDF and htmlzip formats of docs for offline consumption @jaychia (#2538)

🧰 Maintenance

• [CHORE] Remove daft-execution @colin-ho (#2553)
• [CHORE] Enable all features for rust-analyzer @Vince7778 (#2560)
• [CHORE] Add TPC-H questions 11-22 to benchmarks (currently skipped) @kevinzwang (#2299)
• [CHORE]: move minhash to daft-functions @universalmind303 (#2518)

⬆️ Dependencies

• Bump async-compression from 0.4.10 to 0.4.12 @dependabot (#2548)

v0.2.31

Changes

👾 Bug Fixes

• [BUG] Fix bug with map_groups UDFs that return more than 1 output row for empty partitions @jaychia (#2532)
• [BUG] Use shared thread pool for multiple running instances of df on pyrunner @jaychia (#2502)
• [BUG] Fix bug with multi-partition any_value @Vince7778 (#2531)
• [BUG] Allow for Parquet reading from files with differing schemas @jaychia (#2514)
• [BUG] With_new_children not implemented for sample @colin-ho (#2528)

🧰 Maintenance

• [CHORE] add pytest benchmarking for local testing of execution engine @samster25 (#2523)

v0.2.30

Changes

✨ New Features

• [FEAT] Decouple pipeline building and running from new executor @colin-ho (#2522)
• [FEAT] Add concat to new execution model + buffered intermediate ops @colin-ho (#2519)
• [FEAT] Math Ops for FixedSizeList / FixedShapeTensor / Embedding Type @samster25 (#2507)
• [FEAT] adding transform functionality @otacilio-psf (#2498)
• [FEAT] List chunk expression @desmondcheongzx (#2491)
• [FEAT] Refactors and agg improvements for new local execution model @colin-ho (#2497)
• [FEAT]: dyn function registry @universalmind303 (#2466)
• [FEAT] New Local Execution Model @colin-ho (#2437)
• [FEAT] Implement hashing and groupby on lists @Vince7778 (#2464)
• [FEAT] Add upload functionality to binary columns @jaychia (#2461)
• [FEAT] List slice expression @desmondcheongzx (#2479)
• [FEAT] add bit shift functions @murex971 (#2453)
• [FEAT] Implements trigonometry expressions: arctanh arccosh arcsinh @fedemagnani (#2476)
• [FEAT] Microsoft Fabric support in AzureConfig @kevinzwang (#2465)

🚀 Performance Improvements

• [PERF]: dont read parquet metadata multiple times @universalmind303 (#2358)
• [PERF] Local Execution Plan @samster25 (#2489)
• [PERF] Optimize string normalization @Vince7778 (#2474)

👾 Bug Fixes

• [BUG] merge conflict for python catalog scan task @samster25 (#2517)
• [BUG] Add retries for write timeout errors @mauriceweber (#2508)
• [BUG] Use Daft s3 credentials chain for deltalake reads @jaychia (#2486)
• [BUG] Support delta-rs version >0.17.4 in deltalake writes @jaychia (#2488)
• [BUG] Fix anti-join on different column names @Vince7778 (#2477)

📖 Documentation

• [FEAT] adding transform functionality @otacilio-psf (#2498)
• [DOCS] Tutorial: FOTW data access @avriiil (#2384)
• [FEAT] List chunk expression @desmondcheongzx (#2491)
• [DOCS] Enhance Dataframe / Expressions examples @sunaysanghani (#2360)
• [FEAT] add bit shift functions @murex971 (#2453)
• [FEAT] Implements trigonometry expressions: arctanh arccosh arcsinh @fedemagnani (#2476)
• [DOCS] Adds RunLLM widget @vsreekanti (#2462)
• [DOCS] Fix broken example URL @kevinzwang (#2467)
• [FEAT] Microsoft Fabric support in AzureConfig @kevinzwang (#2465)
• [DOCS] Remove unsupported map get doc example @kevinzwang (#2452)

🧰 Maintenance

• [CHORE] remove tokio-stream as dep @samster25 (#2521)
• [CHORE] drop unused deps and add machete CI check @samster25 (#2520)
• [CHORE] fix incremental builds with vscode rust analyzer @samster25 (#2515)
• [CHORE] Disable Python as default feature and have maturin enable it by default @samster25 (#2516)
• [CHORE]: remove daft-core from daft-io @universalmind303 (#2513)
• [CHORE]: move uri functions to new "daft-functions" crate @universalmind303 (#2501)
• [CHORE] Add better typing for class UDFs @jaychia (#2388)
• [CHORE] Add a helpful message in build-artifact-s3 workflow @jaychia (#2480)

v0.2.29

Changes

✨ New Features

• [FEAT] String normalize expression @Vince7778 (#2450)
• [FEAT] Add a IOConfig.http with initial option for user_agent @jaychia (#2449)
• [FEAT] Add Schema.to_pyarrow_schema() @jaychia (#2447)
• [FEAT] thread through transport errors @samster25 (#2446)
• [FEAT] Add MinHash expression @Vince7778 (#2431)
• [FEAT]: write lance @universalmind303 (#2421)
• [FEAT] Add struct get syntactic sugar @kevinzwang (#2367)
• [FEAT] Bitwise 'AND' 'OR' 'XOR' Operations @mrutunjay-kinagi (#2365)
• [FEAT] Add bearer token authentication for Azure @kevinzwang (#2436)
• [FEAT] Add ability to specify snapshot_id for iceberg read @jaychia (#2426)
• [FEAT]: hash expr @universalmind303 (#2398)
• [FEAT] Fixed Size Binary Type v2 @Vince7778 (#2403)
• [FEAT] Automatically use Ray Runner if Ray is initialized @jaychia (#2282)
• [FEAT] print more info when hitting todo in stage planner @samster25 (#2416)
• [FEAT] Add casting from lists to embeddings @Vince7778 (#2396)

👾 Bug Fixes

• [BUG] Use estimated in-memory size for scan task merging and resource requests @kevinzwang (#2448)
• [BUG] Add retries to pyarrow write_dataset call @kevinzwang (#2445)
• [BUG] Fix time type inference @colin-ho (#2441)
• [BUG] Enable display of time64 with seconds unit @jaychia (#2439)
• [BUG] Raise error when Ray Data tensor cannot be pickled and disable compliant nested types @kevinzwang (#2428)
• [BUG] Remove debug print @Vince7778 (#2419)
• [BUG] [New Executor] Drain channel in limit sink @colin-ho (#2401)
• [BUG] Slice array if necessary when casting from list to fixed size list @colin-ho (#2415)

📖 Documentation

• [FEAT] String normalize expression @Vince7778 (#2450)
• [FEAT] Add Schema.to_pyarrow_schema() @jaychia (#2447)
• [FEAT] Add MinHash expression @Vince7778 (#2431)
• [DOCS] Improve groupby and agg docs @kevinzwang (#2438)
• [DOCS] Update unity-catalog docs to include installation @jaychia (#2440)
• [FEAT] Automatically use Ray Runner if Ray is initialized @jaychia (#2282)

v0.2.28

Changes

✨ New Features

• [FEAT] Add manual auth for GCS and Iceberg GCS auth support @kevinzwang (#2393)
• [FEAT] Replace schema_hints with schema and infer_schema for read_json @GuyPozner (#2357)
• [FEAT] [New Executor] [3/N] Full execution model prototype. @clarkzinzow (#2347)
• [FEAT] date and timestamp parsers @murex971 (#2353)
• [FEAT] Implement Anti and Semi Join @samster25 (#2379)
• [FEAT] Implement arctan2 expression @Vince7778 (#2389)
• [FEAT] Add Unity Catalog support @jaychia (#2377)
• [FEAT] fill_nan and not_nan expressions @colin-ho (#2313)
• [FEAT] Add more context when UDFs fail @jaychia (#2325)
• [FEAT] [New Executor] [2/N] daft-execution crate + proof-of-concept compute ops and partition reference + metadata model for new executor. @clarkzinzow (#2340)

👾 Bug Fixes

• [BUG] with_column with existing column name should not reorder columns @colin-ho (#2381)
• [BUG] Fix IO integration tests @jaychia (#2390)
• [BUG]: ide completions for expr namespaces @universalmind303 (#2374)
• [BUG] enable empty stats for deltalake @samster25 (#2376)
• [BUG] Allow variable columns in CSV @colin-ho (#2326)

📖 Documentation

• [FEAT] date and timestamp parsers @murex971 (#2353)
• [FEAT] Implement arctan2 expression @Vince7778 (#2389)
• [FEAT] Add Unity Catalog support @jaychia (#2377)
• [FEAT] fill_nan and not_nan expressions @colin-ho (#2313)
• [EXPRESSIONS] Add missing doc gen for Expression.float.is_nan @tlm365 (#2378)
• [DOCS] add Delta writer functionality @avriiil (#2372)
• [EXPRESSIONS] Implement Expression.float.is_inf @tlm365 (#2371)
• [DOCS] Add tutorial for Data + AI Summit 2024 @jaychia (#2368)
• [DOCS] improve SQL docs @avriiil (#2361)
• [DOCS] the project currently uses ray which is not compatible with python 3.12 @prabodh1194 (#2354)
• [CHORE]: make parquet metadata (de)serializable @universalmind303 (#2346)
• [CHORE]: Arrow2 migrate @universalmind303 (#2341)

🧰 Maintenance

• [CHORE] Run doctests in CI @colin-ho (#2362)
• [CHORE] Bump chrono to 0.4.38 @colin-ho (#2352)
• [CHORE]: make parquet metadata (de)serializable @universalmind303 (#2346)
• [CHORE]: Arrow2 migrate @universalmind303 (#2341)

⬆️ Dependencies

• Bump num-traits from 0.2.18 to 0.2.19 @dependabot (#2285)

v0.2.27

Changes

✨ New Features

• [FEAT] Introduce terminal wrap around when explaining plans @samster25 (#2342)
• [FEAT] [New Executor] [1/N] Move physical plan scheduler to new crate, misc. refactorings + drive-bys for new executor @clarkzinzow (#2339)

👾 Bug Fixes

• [BUG] Azure and Iceberg read and write fixes @kevinzwang (#2349)
• [BUG] Allow nulls in partition column @colin-ho (#2344)
• [BUG] Allow any file extension in Azure directory listing test @kevinzwang (#2345)
• [BUG] Add Retry on Streaming Errors when collecting stream into bytes @samster25 (#2338)

v0.2.26

Changes

✨ New Features

• [FEAT] Public Delta Lake writer @kevinzwang (#2329)
• [FEAT] Implement str.substr Expression @danila-b (#2269)
• [FEAT] Add additional Azure authentication methods @kevinzwang (#2333)
• [FEAT] Custom S3 Credentials Provider @kevinzwang (#2233)
• [FEAT] Expression between @GuyPozner (#2301)
• [FEAT] Allow for printing of plans to a file @jaychia (#2320)

👾 Bug Fixes

• [BUG] Use os.path.join in read_hudi only for local fs @colin-ho (#2336)
• [BUG] Translate mssql to tsql in read_sql scan @colin-ho (#2330)
• [BUG]Fix missing columns after join @siddharth-gulia (#2321)

📖 Documentation

• [FEAT] Public Delta Lake writer @kevinzwang (#2329)
• [FEAT] Implement str.substr Expression @danila-b (#2269)
• [FEAT] Custom S3 Credentials Provider @kevinzwang (#2233)
• [FEAT] Expression between @GuyPozner (#2301)
• [EXPRESSIONS] Add log with custom base @jhasm (#2324)

🧰 Maintenance

• [CHORE] Show full schema in html repr @colin-ho (#2327)

⬆️ Dependencies

• Bump rayon from 1.8.0 to 1.10.0 @dependabot (#2307)

v0.2.25

Changes

✨ New Features

• [FEAT] Add function to refresh logger state for rust @samster25 (#2323)
• [FEAT] implement timeout for parquet reader @samster25 (#2322)
• [FEAT] Use detached named actor for RayRunner @jaychia (#2296)
• [FEAT] like and ilike functions @murex971 (#2283)
• [FEAT] Delta Lake Writer (non-public API) @kevinzwang (#2304)
• [FEAT] extract minute second and time component @murex971 (#2234)
• [FEAT]: tilde expansion @universalmind303 (#2277)
• [FEAT] Map Getter @colin-ho (#2255)
• [FEAT] Test for partition evolution @Fokko (#2084)
• [FEAT] Handle Hudi empty timeline @xushiyan (#2268)

🚀 Performance Improvements

• [PERF]: local json reader @universalmind303 (#2264)

👾 Bug Fixes

• [BUG] Fix multi-output tasks in RayRunner @jaychia (#2291)
• [BUG] Fix red line in Jupyter notebook @kevinzwang (#2267)

📖 Documentation

• [FEAT] like and ilike functions @murex971 (#2283)
• [DOCS] Add iceberg summit tutorial notebook @jaychia (#2281)
• [FEAT] extract minute second and time component @murex971 (#2234)
• [FEAT] Map Getter @colin-ho (#2255)
• [DOCS] Improve docs for external types @kevinzwang (#2274)
• [DOCS] Add quick docs for struct datatype @jaychia (#2256)

🧰 Maintenance

• [CHORE]: Don't use record fields @Fokko (#2306)
• [CHORE] Pin requests to fix docker-py CI issue @jaychia (#2289)
• [CHORE] Clean up string kernel code @kevinzwang (#2276)
• [CHORE] Empty PR to retrigger coverage @samster25 (#2262)
• [CHORE] turn on cov again @samster25 (#2259)
• [CHORE] Enable CI checks for AQE in CI and refactor CI job organization @samster25 (#2258)

⬆️ Dependencies

• Bump async-compression from 0.4.7 to 0.4.10 @dependabot (#2270)
• Bump tokio-util from 0.7.9 to 0.7.11 @dependabot (#2271)

v0.2.24

Changes

✨ New Features

• [FEAT] Allow returning of pyarrow arrays from UDFs @jaychia (#2252)
• [FEAT] Add left, right, and outer joins @kevinzwang (#2166)
• [FEAT] Add rpad and lpad expressions @murex971 (#2157)
• [FEAT] AWS Profile override in S3Config @samster25 (#2243)
• [FEAT] Add unpivot @kevinzwang (#2204)
• [FEAT] Add string repeat functionality @murex971 (#2198)
• [FEAT] Approximate quantile aggregation (pulled into main) @jaychia (#2179)
• [FEAT] pivot @colin-ho (#2183)

🚀 Performance Improvements

• [PERF] Adaptive Query Execution @samster25 (#2176)
• [PERF]: swap out json_deserializer for simd_json @universalmind303 (#2228)
• [PERF] Evaluate only true/false side of if_else if predicate is boolean @colin-ho (#2222)
• [PERF] enable metadata preservation across materialization points @samster25 (#2216)

👾 Bug Fixes

• [BUG] Fix tab completion on expression namespaced accessors @jaychia (#2251)
• [BUG] route abfss to AzureBlob @samster25 (#2244)

📖 Documentation

• [CHORE] Skip demo notebook @jaychia (#2248)
• [FEAT] Add rpad and lpad expressions @murex971 (#2157)
• [DOCS] Add user guide for read_sql @colin-ho (#2226)
• [FEAT] Add unpivot @kevinzwang (#2204)
• [DOCS] Add read_hudi in the api docs @xushiyan (#2225)
• [FEAT] Add string repeat functionality @murex971 (#2198)
• [DOCS] LinkedIn Big Data meetup tutorial @jaychia (#2223)
• [FEAT] Approximate quantile aggregation (pulled into main) @jaychia (#2179)
• [DOCS] Add read_lance docs @jaychia (#2218)
• [FEAT] pivot @colin-ho (#2183)

🧰 Maintenance

• [CHORE] Drop Python 3.7 @samster25 (#2250)
• [CHORE] Improve timestamp repr @colin-ho (#2245)
• [CHORE] Allow multiple group_bys for pivot @colin-ho (#2242)
• [CHORE] Skip demo notebook @jaychia (#2248)
• [CHORE] Return &str for expression name @colin-ho (#2224)
• [CHORE] Mount provision.py for iceberg integration tests @jaychia (#2232)
• [CHORE]: remove trait aliases @universalmind303 (#2229)

⬆️ Dependencies

• Bump serde from 1.0.198 to 1.0.200 @dependabot (#2239)
• Bump csv-async from 1.2.6 to 1.3.0 @dependabot (#2238)