sketches: Probabilistic Data Structures, in Rust
Fast, memory-efficient probabilistic data structures for streaming analytics, cardinality estimation, quantile computation, and sampling.
Python bindings for Rust-based implementations of HyperLogLog, T-Digest, Reservoir Sampling, and more via PyO3.
Features
| Algorithm Category | Implementation | Description |
|---|---|---|
| Cardinality Estimation | HyperLogLog (HLL) | Industry-standard distinct counting with ~1% error |
| HyperLogLog++ | Enhanced HLL with bias correction and sparse mode | |
| CPC Sketch | Most compact serialisation for network transfer | |
| Linear Counter | Optimal for small cardinalities (n < 1000) | |
| Hybrid Counter | Auto-transitions from Linear to HLL | |
| Set Operations | Theta Sketch | Union, intersection, difference with cardinality estimation |
| Sampling | Algorithm R | Basic reservoir sampling for uniform random samples |
| Algorithm A | Optimised reservoir sampling (Vitter, skips items probabilistically) | |
| Weighted Sampling | Probability-proportional reservoir sampling | |
| Stream Sampling | High-throughput sampling with batching | |
| Quantile Estimation | T-Digest | Superior accuracy for extreme quantiles (p95, p99) |
| KLL Sketch | Provable error bounds (~1.65% at k=200) | |
| Frequency Estimation | Count-Min Sketch | Conservative frequency estimation with epsilon-delta guarantees |
| Count Sketch | Unbiased frequency estimation using median | |
| Frequent Items | Top-K heavy hitters with Space-Saving algorithm | |
| Membership Testing | Bloom Filter | Fast membership testing with configurable false positive rate |
| Counting Bloom | Bloom filter with deletion support | |
| Multi-dimensional | Array of Doubles | Tuple sketch for multi-dimensional aggregation |
Mergeable means two independently built sketches can be combined into one that represents the union of both input streams, without access to the original data. This is essential for distributed systems where data is partitioned across nodes -- each node builds a local sketch, then all sketches are merged into a single result.
Install
The package is not yet published to PyPI (the reserved name is rusty-sketches); install from source for now.
git clone https://github.com/tallamjr/sketches.git
cd sketches
pip install .
For an editable install with development dependencies and the maturin build:
pip install -e .[dev]
maturin develop
Quickstart
from sketches import HllSketch
sketch = HllSketch(lg_k=12)
for item in ["apple", "banana", "orange", "apple"]:
sketch.update(item)
print(f"Estimated unique items: {sketch.estimate():.2f}")
See the usage guide for every sketch with runnable examples.
Choosing the right sketch
Different problems call for different sketches. Use this guide to pick the right one for your use case.
| Problem | "How do I know if..." | Small Scale | Large Scale | Distributed / Mergeable |
|---|---|---|---|---|
| Membership | "Is X in the set?" | BloomFilter |
CountingBloomFilter (if deletions needed) |
Yes, union via bitwise OR |
| Cardinality | "How many unique items?" | LinearCounter (n < 1000) |
HllSketch / HllPlusPlusSketch |
Yes, register-wise max |
| Cardinality + Set Ops | "What's the overlap between A and B?" | ThetaSketch |
ThetaSketch |
Yes, union, intersect, difference |
| Compact Cardinality | "Unique count with minimal serialised size?" | CpcSketch |
CpcSketch |
Yes, sketch merging |
| Frequency | "What are the top-K items?" | FrequentStringsSketch |
CountMinSketch / CountSketch |
Yes, entry-wise addition |
| Quantiles | "What's the p99 latency?" | KllSketch (provable bounds) |
TDigest (extreme quantile accuracy) |
Yes, digest merging |
| Sampling | "Give me a random subset" | ReservoirSamplerR |
ReservoirSamplerA (Vitter, skips items) |
Partial, merge samplers |
| Weighted Sampling | "Sample proportional to weight" | WeightedReservoirSampler |
VarOptSketch (Horvitz-Thompson) |
Yes, VarOpt merge |
| Multi-dimensional | "Aggregate multiple metrics per key" | AodSketch |
AodSketch |
Yes, summary merging |
Key trade-offs:
- HLL vs Theta: HLL is more memory-efficient for pure cardinality. Theta supports set operations (union, intersection, difference).
- HLL vs CPC: CPC achieves ~40% smaller serialised size but is more complex. Use CPC when network transfer cost matters.
- Count-Min vs Count Sketch: Count-Min always overestimates (conservative). Count Sketch is unbiased but uses more space.
- KLL vs T-Digest: KLL has provable error bounds (~1.65% at k=200). T-Digest excels at extreme quantiles (p99, p99.9) but bounds are empirical.
- Algorithm R vs A: Both produce uniform samples. Algorithm A skips items probabilistically (Vitter), which is much faster than Algorithm R for large streams.
Performance
Measured on a stabilised harness (the median over independent rounds with a 95%
bootstrap confidence interval), our xxh3-backed default leads Apache DataSketches
across the board. On real TPC-H string columns we beat hand-tuned Apache C++ on
every one of the five shared sketches, and beat the Apache Rust crate on all five
by wider margins. Accuracy stays at parity: HLL, Theta and CPC match Apache
DataSketches within the run-to-run noise band by multi-trial RMSE, all at or below
the 1/sqrt(k) floor.
Speedup over Apache on a real TPC-H column (the c_address field of a customer
table, about 150k distinct strings, generated with
tpchgen-rs), where every bar clears
the parity line:
The same lead in absolute time on that TPC-H column: a single HLL update takes roughly 7.7 ns with this library against 16.7 ns for Apache C++ (lower is better).
What we benchmark against. The references are the official Apache
implementations built and run locally: apache/datasketches-cpp at master
3.2.0-858-g0bab259 (2026-06-19), built with cmake/g++ at C++11, and the
official Apache Rust datasketches crate. Each runner emits one shared CSV
schema over identical datasets, and the C++ runner has a startup self-check that
aborts if its measurement scaffolding is miscompiled. Throughput is the median
over independent rounds with a 95% bootstrap confidence interval; accuracy is
multi-trial RMSE.
See docs/benchmarks.md for the full methodology, the reproduction steps, and the throughput, memory and accuracy plots.
Documentation
- Background: probabilistic data structures
- Usage guide
- Design and architecture
- Benchmarks
- Algorithm deep dive
- Comparison and prior art
Roadmap
Planned but not yet implemented:
- Similarity estimation: MinHash, SimHash, an LSH framework.
- Performance: SIMD-accelerated register updates (the implementation is currently pure scalar Rust).
- Integration: Polars custom expressions and DataFrame operations.
Current priorities: MinHash first (fills the similarity gap), then Polars expressions, then batch operations for the Python bindings.
License
This project is licensed under the MIT License (see pyproject.toml).
1 maintainer
Metadata
1 maintainer
Release files for rusty-sketches 1.0.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| rusty_sketches-1.0.0.tar.gz | 1.3 MB | Details |
Built distributions (wheels)
Total release size: 8.8 MB
Release files / rusty_sketches-1.0.0.tar.gz
| Download URL | rusty_sketches-1.0.0.tar.gz |
|---|---|
| Size | 1.3 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
e4733b6a24669536bc4e7b48d93fbc29c0b4e6f4428fd348cb3f29688eadc5a4
|
|
BLAKE2b-256 checksum How to use checksums |
10c0329a6aeeabc1a4e491389f3d9487ae077289ccda37ee2a297a24ec8f405c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp313-cp313-win_amd64.whl
| Download URL | rusty_sketches-1.0.0-cp313-cp313-win_amd64.whl |
|---|---|
| Size | 464.1 kB |
| Tags | CPython 3.13 Windows x86-64 |
|
SHA-256 checksum How to use checksums |
20d3d013aa1147b4125c990175b1bad3c870522979f1138c5162cbf340f2a063
|
|
BLAKE2b-256 checksum How to use checksums |
9dd49f4505b34519f4866abcede143e7c73c9c98f72fd4e4a0e4388f79cbaebe
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | rusty_sketches-1.0.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 536.1 kB |
| Tags | CPython 3.13 Linux glibc 2.17+ x86-64 |
|
SHA-256 checksum How to use checksums |
a1b86ce65a29f1aa26947c5f0b082074a718d016be9664011197c2013e4a0fee
|
|
BLAKE2b-256 checksum How to use checksums |
adf951ab2b28dc9f4841818c3c7f733b8b3e4f16dcd5f284e2b649eaa06ed00f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
| Download URL | rusty_sketches-1.0.0-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl |
|---|---|
| Size | 510.8 kB |
| Tags | CPython 3.13 Linux glibc 2.17+ ARM64 |
|
SHA-256 checksum How to use checksums |
61c4303e6998eed6264fdf7b7f333fd2aa588455d375db36575250df4d76888d
|
|
BLAKE2b-256 checksum How to use checksums |
cfec980a73f65ca35b5ad3e5d888282c60d54db71cdf83768c1fab6d0f497f96
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp313-cp313-macosx_11_0_arm64.whl
| Download URL | rusty_sketches-1.0.0-cp313-cp313-macosx_11_0_arm64.whl |
|---|---|
| Size | 488.5 kB |
| Tags | CPython 3.13 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
2b96e580ed23bc90948b697621b92e9f1458c2748ad7ae26d572c46ea7d37586
|
|
BLAKE2b-256 checksum How to use checksums |
dad9458cc104971e6a5e7349d574b8f7452efc5989baed74c09c07b95c1dd417
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp313-cp313-macosx_10_12_x86_64.whl
| Download URL | rusty_sketches-1.0.0-cp313-cp313-macosx_10_12_x86_64.whl |
|---|---|
| Size | 508.4 kB |
| Tags | CPython 3.13 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
8d210b08a1247e9cce1b7a37f198d794fc22fc6641cfbbc2ba0c03b6b9bb696f
|
|
BLAKE2b-256 checksum How to use checksums |
8adbfa9b1095f5b09a9b3073057aa41bf2436ac96fa0069d9a38984898cd2dc7
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp312-cp312-win_amd64.whl
| Download URL | rusty_sketches-1.0.0-cp312-cp312-win_amd64.whl |
|---|---|
| Size | 464.7 kB |
| Tags | CPython 3.12 Windows x86-64 |
|
SHA-256 checksum How to use checksums |
b610f5a614030c6682a4a5554a9fbba65cf307fdf0c676d76afc3c24f5d75a8a
|
|
BLAKE2b-256 checksum How to use checksums |
630f8ffe72aad5e142ccefc7753ab96f03e2f72b02e38ec1e4ffe1c7dcbd1cef
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | rusty_sketches-1.0.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 536.5 kB |
| Tags | CPython 3.12 Linux glibc 2.17+ x86-64 |
|
SHA-256 checksum How to use checksums |
577c2277e2b84758a0b597f3399a4ac3b9b2dfa6f4c142f9dbfe8db1f8d49a0f
|
|
BLAKE2b-256 checksum How to use checksums |
3029038b0cb074b9a266718544075a377414166a3d205e96c7503e5ee20db1f2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
| Download URL | rusty_sketches-1.0.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl |
|---|---|
| Size | 511.0 kB |
| Tags | CPython 3.12 Linux glibc 2.17+ ARM64 |
|
SHA-256 checksum How to use checksums |
cc9daf0a18ad07cdcd199bf9e8cf487d639992d16fd9c8b2a16dd22fcd6447da
|
|
BLAKE2b-256 checksum How to use checksums |
a8edfea112562939bd293220bd75ab812645de045666a95eb1a9f45975885459
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp312-cp312-macosx_11_0_arm64.whl
| Download URL | rusty_sketches-1.0.0-cp312-cp312-macosx_11_0_arm64.whl |
|---|---|
| Size | 488.6 kB |
| Tags | CPython 3.12 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
a7fa3773e498d179fec291685b03e4ff94da790c775d24d4b4d67dcf786bf7b9
|
|
BLAKE2b-256 checksum How to use checksums |
daaf761801900c9c7a7d6e7fd3d839932b968ab21b89d2e06f60c52d5302eca2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp312-cp312-macosx_10_12_x86_64.whl
| Download URL | rusty_sketches-1.0.0-cp312-cp312-macosx_10_12_x86_64.whl |
|---|---|
| Size | 508.8 kB |
| Tags | CPython 3.12 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
0305624fdcf5d6b0cb34ebe41605cb3730452405ffb342702d207daec2226f5b
|
|
BLAKE2b-256 checksum How to use checksums |
59c7f46b1589753176557cea0c698f733e5bff527eb265934aa1685b42bea02a
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp311-cp311-win_amd64.whl
| Download URL | rusty_sketches-1.0.0-cp311-cp311-win_amd64.whl |
|---|---|
| Size | 464.2 kB |
| Tags | CPython 3.11 Windows x86-64 |
|
SHA-256 checksum How to use checksums |
8bc73839b2245555e30b025365d57ba42cc05c77f03eaf5302aac872be50cd28
|
|
BLAKE2b-256 checksum How to use checksums |
c679af2769fe4d5da28be3529257d48ffb674a049d481da16aa577615b5f2f2d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
| Download URL | rusty_sketches-1.0.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl |
|---|---|
| Size | 534.2 kB |
| Tags | CPython 3.11 Linux glibc 2.17+ x86-64 |
|
SHA-256 checksum How to use checksums |
822853493caeec317f2a5cd8d8ac68dba281d44508d032282c2fac371bb70030
|
|
BLAKE2b-256 checksum How to use checksums |
e205812a3fcacd050c7a28466e8276943b27220f59c714c642e5676c10c2095b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
| Download URL | rusty_sketches-1.0.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl |
|---|---|
| Size | 509.6 kB |
| Tags | CPython 3.11 Linux glibc 2.17+ ARM64 |
|
SHA-256 checksum How to use checksums |
ffc1069c1e1bc9d1954d73f325504089d2cc1793a05ab4a061fdba6502f92980
|
|
BLAKE2b-256 checksum How to use checksums |
e91a427f2bf5224fb948fb4a64e8793f0b977128e372a87bbb437d4ccd75cd54
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp311-cp311-macosx_11_0_arm64.whl
| Download URL | rusty_sketches-1.0.0-cp311-cp311-macosx_11_0_arm64.whl |
|---|---|
| Size | 492.4 kB |
| Tags | CPython 3.11 macOS 11.0+ ARM64 |
|
SHA-256 checksum How to use checksums |
3471d4cdda11052ed5bda80d8903843121f56ee4717337e41a7319e0dc859e2d
|
|
BLAKE2b-256 checksum How to use checksums |
e12cb91699cf5a163f14ba35fff0b44f0c4895a0b126aa35262efb6c47efa70e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency logRelease files / rusty_sketches-1.0.0-cp311-cp311-macosx_10_12_x86_64.whl
| Download URL | rusty_sketches-1.0.0-cp311-cp311-macosx_10_12_x86_64.whl |
|---|---|
| Size | 511.6 kB |
| Tags | CPython 3.11 macOS 10.12+ x86-64 |
|
SHA-256 checksum How to use checksums |
cecef0221800db41720a54b84cf02d339730de7740d98e869a1471e54a1d3a1f
|
|
BLAKE2b-256 checksum How to use checksums |
42002c7a098b10b9beb9efab5b963c7af63567db7e35e8f1ca4655afb671c646
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.12
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jun 30, 2026.
Transparency log
