User guide

Choose an algorithm for the form and layout of your data, configure a fit, and inspect the returned equation and diagnostics. The guides below cover these operations and the options for continuing or comparing runs.

Task	Guide
Construct a dataset from arrays or a file	Data requirements
Load a benchmark or published measurement	Dataset catalog
Select the source of derivatives	Derivative sources
Select an algorithm and its parameters	Algorithms
Interpret scores and refit candidate terms for comparison	Score conventions
Render a report or individual figures	Visualization
Continue a search or branch from a checkpoint	Checkpoints and resuming
Compare runs across seeds, algorithms or datasets	Batch experiments
Run or discuss a search through a language model	Experimental controller

For a first installation and fit, start with Getting started. Design explains the platform's contracts; the API reference lists signatures and parameters.