Start here

Install KD in a Python environment, then use the first fit to inspect a bundled field, run SGA-PDE and read its result. For a file of your own, data requirements describes inspection, array mappings and supported layouts.

notebook

First fit

Load the data, fit an algorithm, and read the result.

notebook

Use your own data

Two coordinate arrays and one field array, from NumPy to a fitted equation.

Experimental controller

The controller guide covers a first run with kd-agent: install the extra, configure an endpoint, select a dataset and read the report. The controller can also discuss a search interactively.

Further documentation

Use the user guide for algorithm configuration, score interpretation, figures, checkpoints and batch experiments. The research examples include KdV, noisy Burgers, breaking waves and chromatography.

The architecture page contains the complete component diagram, import layers and execution order of a fit. Design introduces the platform and plugin responsibilities behind that diagram.