Skip to content

Report sum stat for numerical columns to DF#7567

Draft
AdamGS wants to merge 1 commit intodevelopfrom
adamg/df-sum-stat-non-bool
Draft

Report sum stat for numerical columns to DF#7567
AdamGS wants to merge 1 commit intodevelopfrom
adamg/df-sum-stat-non-bool

Conversation

@AdamGS
Copy link
Copy Markdown
Contributor

@AdamGS AdamGS commented Apr 20, 2026

Summary

Report the sum stat we have up to DF, which improved a lot of its stats overhead now AND includes the clickbench specialized planning that all other engines have.

@AdamGS
Report sum stat for numerical columns to DF
9e5ad5e 
Signed-off-by: Adam Gutglick <adam@spiraldb.com>
@AdamGS AdamGS added changelog/chore A trivial change ext/datafusion Relates to the DataFusion integration action/benchmark-sql Trigger SQL benchmarks to run on this PR labels Apr 20, 2026
@github-actions github-actions bot removed the action/benchmark-sql Trigger SQL benchmarks to run on this PR label Apr 20, 2026
@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark FineWeb NVMe failed! Check the workflow run for details.

@robert3005
Copy link
Copy Markdown
Contributor

robert3005 commented Apr 20, 2026

Last I checked the sum fix wasn’t released yet

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026 •
edited
Loading

Polar Signals Profiling Results

Latest Run

Status Commit Job Attempt Link
🟢 Done 9e5ad5e 1 Explore Profiling Data

Powered by Polar Signals Cloud

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark TPC-DS SF=1 on NVME failed! Check the workflow run for details.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

Benchmarks: PolarSignals Profiling

Vortex (geomean): 0.982x ➖

datafusion / vortex-file-compressed (0.982x ➖, 0↑ 0↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
polarsignals_q00/datafusion:vortex-file-compressed 118468420 121031000 0.98
polarsignals_q01/datafusion:vortex-file-compressed 358582665 362856479 0.99
polarsignals_q02/datafusion:vortex-file-compressed 21186628 22522810 0.94
polarsignals_q03/datafusion:vortex-file-compressed 359734250 363253586 0.99
polarsignals_q04/datafusion:vortex-file-compressed 10406665 10274259 1.01
polarsignals_q05/datafusion:vortex-file-compressed 14359968 14547937 0.99
polarsignals_q06/datafusion:vortex-file-compressed 17912587 18025863 0.99
polarsignals_q07/datafusion:vortex-file-compressed 13293660 13226191 1.01
polarsignals_q08/datafusion:vortex-file-compressed 435182586 445652121 0.98
polarsignals_q09/datafusion:vortex-file-compressed 9556276 10025623 0.95

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark TPC-H SF=1 on NVME failed! Check the workflow run for details.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

File Sizes: PolarSignals Profiling

No file size changes detected.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark TPC-H SF=1 on S3 failed! Check the workflow run for details.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark TPC-H SF=10 on NVME failed! Check the workflow run for details.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark TPC-H SF=10 on S3 failed! Check the workflow run for details.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

🚨🚨🚨❌❌❌ SQL BENCHMARK FAILED ❌❌❌🚨🚨🚨

Benchmark Clickbench on NVME failed! Check the workflow run for details.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

Benchmarks: FineWeb S3

Verdict: No clear signal (environment too noisy confidence)
Attributed Vortex impact: -3.0%
Vortex (geomean): 1.069x ➖
Parquet (geomean): 1.102x ➖
Shifts: Parquet (control) +10.2% · Median polish +3.4%

datafusion / vortex-file-compressed (0.996x ➖, 0↑ 0↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
fineweb_q00/datafusion:vortex-file-compressed 35813112 33497599 1.07
fineweb_q01/datafusion:vortex-file-compressed 532844063 589178660 0.90
fineweb_q02/datafusion:vortex-file-compressed 523125075 531104788 0.98
fineweb_q03/datafusion:vortex-file-compressed 1418983850 1378377841 1.03
fineweb_q04/datafusion:vortex-file-compressed 1363620202 1362665480 1.00
fineweb_q05/datafusion:vortex-file-compressed 1362267506 1336790007 1.02
fineweb_q06/datafusion:vortex-file-compressed 1461956288 1512824477 0.97
fineweb_q07/datafusion:vortex-file-compressed 1375091391 1332975852 1.03
fineweb_q08/datafusion:vortex-file-compressed 534205911 550641945 0.97
datafusion / vortex-compact (1.094x ➖, 0↑ 1↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
fineweb_q00/datafusion:vortex-compact 69043234 53210627 1.30
fineweb_q01/datafusion:vortex-compact 🚨 794277914 570368216 1.39
fineweb_q02/datafusion:vortex-compact 610267821 599265609 1.02
fineweb_q03/datafusion:vortex-compact 1848396724 1817507958 1.02
fineweb_q04/datafusion:vortex-compact 2284127568 2308222080 0.99
fineweb_q05/datafusion:vortex-compact 1692042391 1694840621 1.00
fineweb_q06/datafusion:vortex-compact 1469931475 1462022088 1.01
fineweb_q07/datafusion:vortex-compact 1424726789 1428020492 1.00
fineweb_q08/datafusion:vortex-compact 467711011 387696936 1.21
datafusion / parquet (1.155x ➖, 0↑ 2↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
fineweb_q00/datafusion:parquet 🚨 1389268978 968615136 1.43
fineweb_q01/datafusion:parquet 2025113750 1646910627 1.23
fineweb_q02/datafusion:parquet 🚨 2357154508 1686814396 1.40
fineweb_q03/datafusion:parquet 1805188601 1601768735 1.13
fineweb_q04/datafusion:parquet 1913549983 1646544131 1.16
fineweb_q05/datafusion:parquet 1932954677 1730823513 1.12
fineweb_q06/datafusion:parquet 1910729587 1839921694 1.04
fineweb_q07/datafusion:parquet 1784836951 1862137668 0.96
fineweb_q08/datafusion:parquet 1907150669 1864417938 1.02
duckdb / vortex-file-compressed (1.110x ➖, 0↑ 1↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
fineweb_q00/duckdb:vortex-file-compressed 🚨 146437299 82390626 1.78
fineweb_q01/duckdb:vortex-file-compressed 587244269 527120311 1.11
fineweb_q02/duckdb:vortex-file-compressed 463552264 442862224 1.05
fineweb_q03/duckdb:vortex-file-compressed 1508383234 1521152413 0.99
fineweb_q04/duckdb:vortex-file-compressed 1620634129 1621734191 1.00
fineweb_q05/duckdb:vortex-file-compressed 1554292014 1505090521 1.03
fineweb_q06/duckdb:vortex-file-compressed 1655842612 1686360061 0.98
fineweb_q07/duckdb:vortex-file-compressed 1608967716 1472967569 1.09
fineweb_q08/duckdb:vortex-file-compressed 711283195 630468417 1.13
duckdb / vortex-compact (1.079x ➖, 0↑ 1↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
fineweb_q00/duckdb:vortex-compact 🚨 124564228 78721832 1.58
fineweb_q01/duckdb:vortex-compact 546432848 544846961 1.00
fineweb_q02/duckdb:vortex-compact 712417684 557495238 1.28
fineweb_q03/duckdb:vortex-compact 1590974621 1694551329 0.94
fineweb_q04/duckdb:vortex-compact 1762496763 1808923811 0.97
fineweb_q05/duckdb:vortex-compact 1546152793 1628015096 0.95
fineweb_q06/duckdb:vortex-compact 1557847535 1475388586 1.06
fineweb_q07/duckdb:vortex-compact 1392996588 1299736887 1.07
fineweb_q08/duckdb:vortex-compact 469341771 472218049 0.99
duckdb / parquet (1.052x ➖, 0↑ 0↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
fineweb_q00/duckdb:parquet 1098359900 1083729914 1.01
fineweb_q01/duckdb:parquet 1458167607 1341198688 1.09
fineweb_q02/duckdb:parquet 1427047082 1473831400 0.97
fineweb_q03/duckdb:parquet 4031094394 3682823969 1.09
fineweb_q04/duckdb:parquet 2188531437 1944651281 1.13
fineweb_q05/duckdb:parquet 2186258127 2204345048 0.99
fineweb_q06/duckdb:parquet 4356714425 4175677131 1.04
fineweb_q07/duckdb:parquet 2654858703 2571892121 1.03
fineweb_q08/duckdb:parquet 1216662542 1084227919 1.12
Full attributed analysis
Query Config Raw Δ Control Δ Attributed α Noise floor Significant?
0 datafusion:vortex-compact +29.8% +20.6% +7.6% +134.7% ➖ noise
0 datafusion:vortex-file-compressed +6.9% +20.6% -11.3% +276.6% ➖ noise
0 duckdb:vortex-compact +58.2% +20.6% +31.2% +68.8% ➖ noise
0 duckdb:vortex-file-compressed +77.7% +20.6% +47.4% +73.8% ➖ noise
1 datafusion:vortex-compact +39.3% +15.6% +20.4% +76.3% ➖ noise
1 datafusion:vortex-file-compressed -9.6% +15.6% -21.8% +76.7% ➖ noise
1 duckdb:vortex-compact +0.3% +15.6% -13.3% +30.0% ➖ noise
1 duckdb:vortex-file-compressed +11.4% +15.6% -3.6% +43.7% ➖ noise
2 datafusion:vortex-compact +1.8% +16.3% -12.5% +30.0% ➖ noise
2 datafusion:vortex-file-compressed -1.5% +16.3% -15.3% +30.0% ➖ noise
2 duckdb:vortex-compact +27.8% +16.3% +9.9% +30.0% ➖ noise
2 duckdb:vortex-file-compressed +4.7% +16.3% -10.0% +30.0% ➖ noise
3 datafusion:vortex-compact +1.7% +11.1% -8.4% +30.0% ➖ noise
3 datafusion:vortex-file-compressed +2.9% +11.1% -7.3% +30.0% ➖ noise
3 duckdb:vortex-compact -6.1% +11.1% -15.5% +30.0% ➖ noise
3 duckdb:vortex-file-compressed -0.8% +11.1% -10.7% +30.0% ➖ noise
4 datafusion:vortex-compact -1.0% +14.4% -13.5% +30.0% ➖ noise
4 datafusion:vortex-file-compressed +0.1% +14.4% -12.5% +30.0% ➖ noise
4 duckdb:vortex-compact -2.6% +14.4% -14.8% +30.0% ➖ noise
4 duckdb:vortex-file-compressed -0.1% +14.4% -12.6% +30.0% ➖ noise
5 datafusion:vortex-compact -0.2% +5.2% -5.1% +30.0% ➖ noise
5 datafusion:vortex-file-compressed +1.9% +5.2% -3.2% +30.0% ➖ noise
5 duckdb:vortex-compact -5.0% +5.2% -9.8% +30.0% ➖ noise
5 duckdb:vortex-file-compressed +3.3% +5.2% -1.9% +30.0% ➖ noise
6 datafusion:vortex-compact +0.5% +4.1% -3.4% +30.0% ➖ noise
6 datafusion:vortex-file-compressed -3.4% +4.1% -7.2% +30.0% ➖ noise
6 duckdb:vortex-compact +5.6% +4.1% +1.4% +30.0% ➖ noise
6 duckdb:vortex-file-compressed -1.8% +4.1% -5.7% +30.0% ➖ noise
7 datafusion:vortex-compact -0.2% -0.5% +0.3% +30.0% ➖ noise
7 datafusion:vortex-file-compressed +3.2% -0.5% +3.7% +30.0% ➖ noise
7 duckdb:vortex-compact +7.2% -0.5% +7.7% +30.0% ➖ noise
7 duckdb:vortex-file-compressed +9.2% -0.5% +9.8% +30.0% ➖ noise
8 datafusion:vortex-compact +20.6% +7.1% +12.6% +30.0% ➖ noise
8 datafusion:vortex-file-compressed -3.0% +7.1% -9.4% +37.1% ➖ noise
8 duckdb:vortex-compact -0.6% +7.1% -7.2% +35.6% ➖ noise
8 duckdb:vortex-file-compressed +12.8% +7.1% +5.3% +30.0% ➖ noise

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

Benchmarks: Statistical and Population Genetics

Verdict: No clear signal (low confidence)
Attributed Vortex impact: +1.5%
Vortex (geomean): 0.985x ➖
Parquet (geomean): 0.971x ➖
Shifts: Parquet (control) -2.9% · Median polish -1.7%

duckdb / vortex-file-compressed (0.978x ➖, 0↑ 0↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
statpopgen_q00/duckdb:vortex-file-compressed 11392926 12248184 0.93
statpopgen_q01/duckdb:vortex-file-compressed 20404868 20103561 1.01
statpopgen_q02/duckdb:vortex-file-compressed 1380987335 1400487793 0.99
statpopgen_q03/duckdb:vortex-file-compressed 3087588552 3140401947 0.98
statpopgen_q04/duckdb:vortex-file-compressed 3090981355 3207817932 0.96
statpopgen_q05/duckdb:vortex-file-compressed 1459033047 1475593576 0.99
statpopgen_q06/duckdb:vortex-file-compressed 2125812113 2180804120 0.97
statpopgen_q07/duckdb:vortex-file-compressed 226272473 229580301 0.99
statpopgen_q08/duckdb:vortex-file-compressed 272224356 275764503 0.99
statpopgen_q09/duckdb:vortex-file-compressed 2934843633 3066772739 0.96
statpopgen_q10/duckdb:vortex-file-compressed 4733556600 4771925316 0.99
duckdb / vortex-compact (0.993x ➖, 0↑ 0↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
statpopgen_q00/duckdb:vortex-compact 11813151 11760022 1.00
statpopgen_q01/duckdb:vortex-compact 166586092 169493916 0.98
statpopgen_q02/duckdb:vortex-compact 1790983027 1830358510 0.98
statpopgen_q03/duckdb:vortex-compact 3495829463 3519686570 0.99
statpopgen_q04/duckdb:vortex-compact 3533104923 3563398574 0.99
statpopgen_q05/duckdb:vortex-compact 1828037949 1859243845 0.98
statpopgen_q06/duckdb:vortex-compact 2667766516 2684645911 0.99
statpopgen_q07/duckdb:vortex-compact 911223977 914906882 1.00
statpopgen_q08/duckdb:vortex-compact 940304832 948221564 0.99
statpopgen_q09/duckdb:vortex-compact 3354080681 3314029489 1.01
statpopgen_q10/duckdb:vortex-compact 5363175581 5405603608 0.99
duckdb / parquet (0.971x ➖, 0↑ 0↓)
name PR 9e5ad5e (ns) base d0a6dba (ns) ratio (PR/base)
statpopgen_q00/duckdb:parquet 320026888 335809782 0.95
statpopgen_q01/duckdb:parquet 410757736 416138813 0.99
statpopgen_q02/duckdb:parquet 996625378 1023779677 0.97
statpopgen_q03/duckdb:parquet 1488748567 1526552479 0.98
statpopgen_q04/duckdb:parquet 1488893478 1552382906 0.96
statpopgen_q05/duckdb:parquet 1000447508 1034672836 0.97
statpopgen_q06/duckdb:parquet 1495536567 1497173486 1.00
statpopgen_q07/duckdb:parquet 1289412506 1310728373 0.98
statpopgen_q08/duckdb:parquet 1284779181 1318831929 0.97
statpopgen_q09/duckdb:parquet 1329664415 1448807697 0.92
statpopgen_q10/duckdb:parquet 2648249525 2676786615 0.99
Full attributed analysis
Query Config Raw Δ Control Δ Attributed α Noise floor Significant?
0 duckdb:vortex-compact +0.5% -4.7% +5.4% +10.0% ➖ noise
0 duckdb:vortex-file-compressed -7.0% -4.7% -2.4% +11.4% ➖ noise
1 duckdb:vortex-compact -1.7% -1.3% -0.4% +25.1% ➖ noise
1 duckdb:vortex-file-compressed +1.5% -1.3% +2.8% +256.9% ➖ noise
2 duckdb:vortex-compact -2.2% -2.7% +0.5% +10.0% ➖ noise
2 duckdb:vortex-file-compressed -1.4% -2.7% +1.3% +10.0% ➖ noise
3 duckdb:vortex-compact -0.7% -2.5% +1.8% +10.0% ➖ noise
3 duckdb:vortex-file-compressed -1.7% -2.5% +0.8% +10.0% ➖ noise
4 duckdb:vortex-compact -0.9% -4.1% +3.4% +10.0% ➖ noise
4 duckdb:vortex-file-compressed -3.6% -4.1% +0.5% +10.0% ➖ noise
5 duckdb:vortex-compact -1.7% -3.3% +1.7% +10.0% ➖ noise
5 duckdb:vortex-file-compressed -1.1% -3.3% +2.3% +10.0% ➖ noise
6 duckdb:vortex-compact -0.6% -0.1% -0.5% +10.0% ➖ noise
6 duckdb:vortex-file-compressed -2.5% -0.1% -2.4% +10.0% ➖ noise
7 duckdb:vortex-compact -0.4% -1.6% +1.2% +10.0% ➖ noise
7 duckdb:vortex-file-compressed -1.4% -1.6% +0.2% +10.0% ➖ noise
8 duckdb:vortex-compact -0.8% -2.6% +1.8% +10.0% ➖ noise
8 duckdb:vortex-file-compressed -1.3% -2.6% +1.3% +10.0% ➖ noise
9 duckdb:vortex-compact +1.2% -8.2% +10.3% +10.0% 🚨 regression
9 duckdb:vortex-file-compressed -4.3% -8.2% +4.3% +10.0% ➖ noise
10 duckdb:vortex-compact -0.8% -1.1% +0.3% +10.0% ➖ noise
10 duckdb:vortex-file-compressed -0.8% -1.1% +0.3% +10.0% ➖ noise

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Apr 20, 2026

File Sizes: Statistical and Population Genetics

No file size changes detected.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

changelog/chore A trivial change ext/datafusion Relates to the DataFusion integration

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@AdamGS @robert3005