Time-series systems research

Time Series Compression Leaderboard

Compare measured storage, encoding, and decoding costs across real time-series datasets.

Loading benchmark snapshot…

…datasets with results

…method families

…implementation runtimes

…runtime / codec configurations

Runtime Data type Float precision Dataset Method rows Method family Find a method

Compression rate = compressed bytes / original bytes. Lower rates and lower ns/point are better.

Only methods covering the full selection

Display columns

Loading measured results…

Frozen paper Table 3 snapshot

These stored C++ results reproduce the project's frozen paper table. They are separate from the current report-tree leaderboard because source reports and method labels have evolved.

Paper data scope

Method	Stored rank	Average rate ↓	Encode ns/point ↓	Decode ns/point ↓	Stored balanced score ↓

Stored rank and balanced score retain the source table's aggregation. The current leaderboard does not use these values to fill missing measurements.

About this benchmark

This Space publishes a snapshot of the existing web_compression project results. The layout follows the grouped leaderboard approach of GIFT-Eval.

Metric definitions

Average compression rate ↓: Arithmetic mean of the per-dataset compression rates in the selected scope.
Overall compression rate ↓: Total compressed bytes divided by total original bytes over the selected measured records.
Encoding / decoding cost ↓: Nanoseconds per data point. The aggregation boundary is documented with the snapshot below.
Coverage: Datasets with usable measurements for a method out of the datasets in the selected scope. Missing results are never filled with zero.

Reading the rankings

Methods with different coverage are measured on different inputs. Coverage counts datasets, not identical columns or point counts. Use the full-coverage filter and inspect a single dataset's input sizes before drawing a direct comparison. Runtime filters keep Python, C++, and Java measurements separate.

By default, each family uses its measured configuration with the lowest compression rate in each dataset. The chosen configuration can vary by dataset. Switch Method rows to Exact configuration to compare a fixed configuration; method details show the exact variant used.

These are published measurements, not experiments executed inside this Space. A reported ratio alone does not establish lossless reconstruction; round-trip checks and hardware details are shown only when their evidence exists in the source snapshot.

Download and provenance

Result snapshot (JSON) · Source manifest (JSON)

Contribute a result

Run the project benchmark and provide the codec/configuration, input identity, encoding and decoding measurements, execution environment, and round-trip validation evidence. Results can be updated from the source project using the snapshot exporter included in the Space repository.