vortex-data/vortex - CodSpeed

vortex

Blog Docs Changelog

Performance History

Latest Results

fix(python): translate Polars is_null expressions (#10189) Polars `is_null()` fails predicate conversion despite native Vortex support. Map the Boolean.IsNull function directly to `ve.is_null`, including null input rows. This is a direct expression translation; it performs no eager evaluation and does not require literal arguments. Literal-specific PRs decode the scalar representation itself. Validation: 1 targeted regression case(s) passed with this PR’s converter. With only `polars_.py` restored to the base revision and the same tests retained, all 1 failed with Polars `ComputeError` wrapping the unsupported-expression `ValueError` or `NotImplementedError`. The tests compare complete dataframes from Polars alone and a Vortex-backed Polars scan, and check explicit row expectations. No lint or broader suite was run. Signed-off-by: Daniel King <dan@spiraldb.com>

develop

1 minute ago

fix(python): translate Polars is_null expressions Signed-off-by: Daniel King <dan@spiraldb.com>

dk/polars-is-null

18 minutes ago

fix(python): translate Polars binary scalar literals (#10188) Polars serializes binary scalar literals as integer byte arrays. Decode those bytes and create a native Vortex binary literal; passing the JSON integer list as the literal value is incorrect. This is a direct expression translation; it performs no eager evaluation and does not require literal arguments. Literal-specific PRs decode the scalar representation itself. Validation: 1 targeted regression case(s) passed with this PR’s converter. With only `polars_.py` restored to the base revision and the same tests retained, all 1 failed with Polars `ComputeError` wrapping the unsupported-expression `ValueError` or `NotImplementedError`. The tests compare complete dataframes from Polars alone and a Vortex-backed Polars scan, and check explicit row expectations. No lint or broader suite was run. Signed-off-by: Daniel King <dan@spiraldb.com>

develop

20 minutes ago

fix(python): translate Polars binary scalar literals Signed-off-by: Daniel King <dan@spiraldb.com>

dk/polars-binary-literals

35 minutes ago

fix(python): translate Polars date literals (#10185) Polars date comparisons fail because `Scalar.Date` is unrecognized. Translate its epoch-day integer into a native Vortex day-resolution date literal, without converting through Python datetime arithmetic. This is a direct expression translation; it performs no eager evaluation and does not require literal arguments. Literal-specific PRs decode the scalar representation itself. Validation: 1 targeted regression case(s) passed with this PR’s converter. With only `polars_.py` restored to the base revision and the same tests retained, all 1 failed with Polars `ComputeError` wrapping the unsupported-expression `ValueError` or `NotImplementedError`. The tests compare complete dataframes from Polars alone and a Vortex-backed Polars scan, and check explicit row expectations. No lint or broader suite was run. --------- Signed-off-by: Daniel King <dan@spiraldb.com>

develop

35 minutes ago

perf(layout): reduce repeated work in sparse V2 scans Compose small identity row-index scans into a projection graph that follows each column's chunks, avoiding repeated decoding at other columns' cuts. Coarsen outer splits only above max(8, 2 * workers) active tasks. Preserve natural granularity for small scans even on one CPU, and retain the existing projection path for contiguous per-task selections. Reuse populated bounds discovered during IO announcements to trim empty edges without rediscovering populated cuts. Coalesce unselected chunk gaps and remove empty arrays before joining nested fields. Keep decoded-segment caching disabled and preserve existing dictionary and segment cache behavior. Add regression coverage for exact read counts across repeated executions, row ranges and limits; empty coverage and array identity; active task counts; and the row bounds handed to projection morsels. Validation: 601 vortex-layout library tests, V1 comparison on boundary selection sizes, and non-compact random-access benchmark comparisons. Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

ji/planning-plan-planexec

36 minutes ago

fix(python): reconcile date literals with typed scalars after rebase Signed-off-by: Claude <noreply@anthropic.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01E92wUoS3LaqWVJmmS7vKVq

dk/polars-date-literals

52 minutes ago

docs(bench): record validated DataFusion V2 integration Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

ji/planning-plan-planexec

53 minutes ago

Latest Branches

-15%

fix(python): translate Polars is_null expressions#10189

6 hours ago

61b090a

dk/polars-is-null

-11%

fix(python): translate Polars binary scalar literals#10188

6 hours ago

d6cf0f4

dk/polars-binary-literals

+39%

do not merge(scan): planning, physical plans, and plan execution#10339

48 minutes ago

6115639

ji/planning-plan-planexec

© 2026 CodSpeed Technology

Home Terms Privacy Docs