Refactor DictEncoder to not use ArrayAccessors#7759
Merged
Conversation
robert3005
added
action/benchmark
github-actions
Bot
removed
the
action/benchmark
Merging this PR will improve performance by 12.63%
|
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ❌ | Simulation | chunked_varbinview_opt_canonical_into[(1000, 10)] |
169.8 µs | 205.9 µs | -17.56% |
| ❌ | Simulation | chunked_varbinview_opt_into_canonical[(1000, 10)] |
183.3 µs | 219.1 µs | -16.36% |
| ❌ | Simulation | slice_empty_vortex |
310 ns | 368.3 ns | -15.84% |
| ⚡ | Simulation | encode_varbin[(1000, 2)] |
285 µs | 145.7 µs | +95.63% |
| ⚡ | Simulation | chunked_bool_canonical_into[(1000, 10)] |
26.7 µs | 16.1 µs | +65.65% |
| ⚡ | Simulation | chunked_varbinview_into_canonical[(100, 100)] |
306.8 µs | 270.6 µs | +13.41% |
| ⚡ | Simulation | encode_primitives[u8, (10000, 512)] |
340.2 µs | 302.2 µs | +12.56% |
| ⚡ | Simulation | encode_primitives[u8, (10000, 4)] |
322.1 µs | 287.2 µs | +12.15% |
| ⚡ | Simulation | encode_primitives[u8, (10000, 2)] |
322 µs | 287.2 µs | +12.13% |
| ⚡ | Simulation | encode_primitives[u8, (10000, 32)] |
325 µs | 290.2 µs | +12.01% |
| ⚡ | Simulation | encode_primitives[u8, (10000, 8)] |
322.7 µs | 288.2 µs | +11.98% |
| ⚡ | Simulation | bitwise_not_vortex_buffer_mut[128] |
273.6 ns | 244.4 ns | +11.93% |
| ⚡ | Simulation | eq_i64_constant |
319.1 µs | 288.2 µs | +10.69% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing rk/cardinality-estimator (0e10063) with develop (9382028)
Footnotes
-
4 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩
Contributor
Polar Signals Profiling ResultsLatest Run
Previous Runs (10)
Powered by Polar Signals Cloud |
Contributor
Benchmarks: CompressionVortex (geomean): 0.993x ➖ How to read Verdict and Engines
unknown / unknown (0.998x ➖, 4↑ 4↓)
|
Contributor
Benchmarks: PolarSignals ProfilingVortex (geomean): 0.991x ➖ How to read Verdict and Engines
datafusion / vortex-file-compressed (0.991x ➖, 1↑ 0↓)
File Size Changes (1 files changed, +0.0% overall, 1↑ 0↓)
Totals:
|
Contributor
Benchmarks: FineWeb NVMeVerdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (1.013x ➖, 0↑ 1↓)
datafusion / vortex-compact (0.989x ➖, 0↑ 0↓)
datafusion / parquet (1.000x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed (0.984x ➖, 1↑ 0↓)
duckdb / vortex-compact (1.006x ➖, 0↑ 0↓)
duckdb / parquet (1.016x ➖, 0↑ 0↓)
File Size Changes (1 files changed, +0.0% overall, 1↑ 0↓)
Totals:
|
Contributor
File Sizes: FineWeb NVMeNo file size changes detected. |
Contributor
Benchmarks: TPC-DS SF=1 on NVMEVerdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (0.991x ➖, 1↑ 0↓)
datafusion / vortex-compact (0.994x ➖, 1↑ 2↓)
datafusion / parquet (1.000x ➖, 0↑ 1↓)
duckdb / vortex-file-compressed (0.993x ➖, 2↑ 0↓)
duckdb / vortex-compact (0.997x ➖, 0↑ 2↓)
duckdb / parquet (0.997x ➖, 0↑ 0↓)
duckdb / duckdb (1.001x ➖, 2↑ 2↓)
File Size Changes (18 files changed, -0.0% overall, 12↑ 6↓)
Totals:
|
Contributor
Benchmarks: Statistical and Population GeneticsVerdict: No clear signal (low confidence) How to read Verdict and Engines
duckdb / vortex-file-compressed (0.990x ➖, 0↑ 0↓)
duckdb / vortex-compact (0.996x ➖, 0↑ 0↓)
duckdb / parquet (1.007x ➖, 0↑ 0↓)
File Size Changes (2 files changed, +0.0% overall, 1↑ 1↓)
Totals:
|
Contributor
Benchmarks: FineWeb S3Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (1.408x ❌, 0↑ 5↓)
datafusion / vortex-compact (1.281x ➖, 0↑ 3↓)
datafusion / parquet (1.384x ❌, 0↑ 8↓)
duckdb / vortex-file-compressed (1.042x ➖, 0↑ 0↓)
duckdb / vortex-compact (1.038x ➖, 0↑ 0↓)
duckdb / parquet (1.096x ➖, 0↑ 0↓)
|
Contributor
Benchmarks: Clickbench on NVMEVerdict: No clear signal (environment too noisy confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (1.044x ➖, 2↑ 7↓)
datafusion / parquet (1.041x ➖, 1↑ 2↓)
duckdb / vortex-file-compressed (1.045x ➖, 0↑ 4↓)
duckdb / parquet (1.007x ➖, 1↑ 1↓)
duckdb / duckdb (1.022x ➖, 1↑ 1↓)
File Size Changes (189 files changed, +1.0% overall, 71↑ 118↓)
Totals:
|
Contributor
File Sizes: Clickbench on NVMEFile Size Changes (105 files changed, +0.0% overall, 95↑ 10↓)
Totals:
|
Contributor
Benchmarks: TPC-H SF=1 on NVMEVerdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (1.087x ➖, 0↑ 7↓)
datafusion / vortex-compact (1.083x ➖, 0↑ 7↓)
datafusion / parquet (1.055x ➖, 0↑ 3↓)
datafusion / arrow (1.107x ❌, 0↑ 11↓)
duckdb / vortex-file-compressed (1.074x ➖, 0↑ 4↓)
duckdb / vortex-compact (1.062x ➖, 0↑ 3↓)
duckdb / parquet (1.047x ➖, 0↑ 4↓)
duckdb / duckdb (1.045x ➖, 0↑ 0↓)
File Size Changes (12 files changed, +0.0% overall, 5↑ 7↓)
Totals:
|
Contributor
File Sizes: TPC-H SF=1 on NVMENo file size changes detected. |
Contributor
Benchmarks: TPC-H SF=10 on NVMEVerdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (0.998x ➖, 0↑ 0↓)
datafusion / vortex-compact (0.994x ➖, 0↑ 0↓)
datafusion / parquet (0.997x ➖, 0↑ 0↓)
datafusion / arrow (1.001x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed (0.992x ➖, 0↑ 0↓)
duckdb / vortex-compact (0.997x ➖, 0↑ 0↓)
duckdb / parquet (0.995x ➖, 0↑ 0↓)
duckdb / duckdb (0.992x ➖, 0↑ 0↓)
File Size Changes (40 files changed, +0.0% overall, 14↑ 26↓)
Totals:
|
Contributor
File Sizes: TPC-H SF=10 on NVMEFile Size Changes (2 files changed, +0.0% overall, 2↑ 0↓)
Totals:
|
Contributor
Benchmarks: TPC-H SF=1 on S3Verdict: No clear signal (environment too noisy confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (1.540x ❌, 0↑ 17↓)
datafusion / vortex-compact (1.435x ❌, 0↑ 14↓)
datafusion / parquet (1.321x ❌, 0↑ 11↓)
duckdb / vortex-file-compressed (1.064x ➖, 0↑ 0↓)
duckdb / vortex-compact (1.046x ➖, 0↑ 0↓)
duckdb / parquet (1.042x ➖, 0↑ 1↓)
|
Contributor
Benchmarks: TPC-H SF=10 on S3Verdict: No clear signal (environment too noisy confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed (1.281x ➖, 0↑ 10↓)
datafusion / vortex-compact (1.191x ➖, 0↑ 3↓)
datafusion / parquet (1.312x ❌, 0↑ 11↓)
duckdb / vortex-file-compressed (1.136x ➖, 0↑ 1↓)
duckdb / vortex-compact (1.120x ➖, 0↑ 0↓)
duckdb / parquet (1.156x ➖, 0↑ 3↓)
|
github-actions
Bot
removed
the
action/benchmark
Contributor
File Sizes: PolarSignals ProfilingNo file size changes detected. |
robert3005
force-pushed
the
rk/cardinality-estimator
branch
from
fc38e36 to
7fa9309
Compare
Contributor
File Sizes: TPC-DS SF=1 on NVMEFile Size Changes (4 files changed, +0.0% overall, 2↑ 2↓)
Totals:
|
github-actions
Bot
removed
the
action/benchmark
Contributor
File Sizes: Statistical and Population GeneticsFile Size Changes (2 files changed, +0.0% overall, 1↑ 1↓)
Totals:
|
Contributor
Benchmarks: Random AccessVortex (geomean): 0.898x ✅ How to read Verdict and Engines
unknown / unknown (0.990x ➖, 12↑ 4↓)
|
Contributor
Author
|
Need to figure out what to do about WASM here, cloudflare crate requires 64 bit pointers |
Contributor
Author
|
I gave up on trying to use cardinality estimator for dict compression. I have kept whatever was useful from that pr and maybe will attempt again in the future |
myrrc
approved these changes
Jun 19, 2026
robert3005
force-pushed
the
rk/cardinality-estimator
branch
from
a936fd4 to
90ca40a
Compare
robert3005
force-pushed
the
rk/cardinality-estimator
branch
from
7f744e5 to
a7fe491
Compare
Replace the exact `HashMap`/`HashSet` previously used to compute distinct-value counts during compression stats generation with Cloudflare's `cardinality-estimator` crate. The estimator gives us a bounded-memory approximation (exact up to ~128 distinct values, then HyperLogLog++) so high-cardinality arrays no longer require an O(n) auxiliary hash table to answer the single question "how many unique values does this have?". - Integer stats swap the hash map for a `CardinalityEstimator` and track the most frequent value via a Boyer-Moore majority candidate plus a second-pass exact count. Sparse/dict schemes only care about the heavy hitter (>= 90% threshold) or a rough distinct ratio, so this is behaviourally equivalent for the decisions they make. - Float and string stats likewise drop their hash sets in favor of the estimator. - The integer and float dictionary encoders now rebuild the exact set of distinct values from the source array at compress time, since they need the values themselves and the stats layer no longer retains them. - `SequenceScheme`'s fast-path check for "all values are distinct" now tolerates the estimator's small approximation error; the deferred callback still validates sequences exactly. Signed-off-by: Robert Kruszewski <github@robertk.io>
robert3005
force-pushed
the
rk/cardinality-estimator
branch
from
1adfb50 to
0e10063
Compare
AdamGS
reviewed
Jun 23, 2026
| mod tests { | ||
| use std::sync::LazyLock; | ||
|
|
||
| #[expect(unused_imports)] |
AdamGS
approved these changes
Jun 23, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
ArrayAccessor is not the fastest api and we can do better by accessing the value iterators directly