Crate touchstone_rs 

Touchstone — a streaming anomaly-detection benchmark framework.

Implement the Detector trait, register it with Touchstone, and call Touchstone::run to evaluate it against a directory of CSV datasets. Results are returned as a polars::prelude::DataFrame with one row per dataset × detector and one column per metric.

Modules

loader

metrics

Macros

touchstone_main

Generates a fn main that runs the given Detector.

Structs

RunArgs

Command-line arguments shared by every algorithm binary.

Touchstone

Evaluation runner for streaming detectors on Touchstone datasets.

Traits

Detector

Streaming detector interface used by Touchstone during evaluation.

DetectorFactory

Factory for creating fresh detector instances with a specific dimensionality.

Functions

run_cli

Parses CLI args, runs a Touchstone evaluation for detector D, and prints the report. The display name is taken from D::name().

Type Aliases

Result

Result<T, Error>