DynaHist: A Dynamic Histogram Library for Java
TypeScript implementation of the distributed quantile sketch algorithm DDSketch
C++14 port of the DDSketch distributed quantile sketch algorithm
Kernel density integral transformation: feature preprocessing and univariate clustering (TMLR, 2023)
R Package for Multivariate Ranks and Quantiles Using Optimal Transportation
O(k) space unbiased stream reservoir sampling and quantile estimation engine for LLM telemetry and high-volume token streams
O(k) space unbiased stream reservoir sampling and quantile estimation engine for LLM telemetry and high-volume token streams
High-integrity Python KLL sketch for streaming quantiles — mergeable, deterministic, zero-dependency, with optional C++17/SIMD acceleration.
Probabilistic data structures for JavaScript and TypeScript: Bloom filter (scalable, blocked), Cuckoo filter, Binary Fuse filter (an xor filter), HyperLogLog, Theta sketch, Count-Min sketch, Top-K, KLL sketch, DDSketch quantile sketch, MinHash and RIBLT set reconciliation. Zero dependencies.
Probabilistic demand forecasting for Ruby: Croston, SBA and TSB for intermittent demand, demand classification with automatic model selection, and calibrated quantiles for every period and the lead-time total. Pure Ruby, zero runtime dependencies.
Inspect and decode Apache DataSketches binaries. Find a serialized sketch in MongoDB, Parquet, Delta, Spark or Databricks, and get a readable report in the browser or CLI
TimesFM 2.5 fork that restores forecast_with_covariates and maps the quantile head to inventory policy.
Python/Seaborn script drawing letter-value (boxen) plots of depth distributions, demonstrated on Mariana Trench bathymetry. Reusable scientific-plotting example.
A Node-Red node realizing the 𝗣² algorithm of Jain and Chlamtec (1985) for dynamic calculation of quantiles without storing observations
Interactive probability-space axis and tick-rendering experiment for adaptive quantitative visualization.
Interactive R lab showing why quantiles cannot be averaged and how mergeable histograms preserve tail latency.
To associate your repository with the quantiles topic, visit your repo's landing page and select "manage topics."