Skip to content

Usage

This section describes how to use PROTEUS. The framework can be run three ways:

  • Standalone: a single forward model from one configuration file.
  • As a grid (ensemble): many forward models sweeping over parameters.
  • Within a retrieval framework: PROTEUS as the forward model in a parameter inference loop.

In all cases you configure the model through a configuration file, described in the configuration guide. If you run into problems, see the troubleshooting page.

Quick start

proteus start -c input/all_options.toml
Results appear in a timestamped output/run_<timestamp>_xxxx/ folder (this config sets params.out.path = "auto"). See Running and output for details.

Where to go next

Page What it covers
Running and output Launching a single run from the terminal, running on remote machines, where results are written, and archiving output.
Initial thermal conditions Setting the mantle's starting temperature profile, what the initial state controls, and which option to favour.
Parameter grids Defining and dispatching ensembles of simulations, with or without Slurm.
Postprocessing and chemistry Atmospheric chemistry with VULCAN, synthetic observations, and multi-angle thermal profiles.
Bayesian inference Using PROTEUS as the forward model in a Bayesian-optimisation retrieval.

Related pages: the configuration file reference, diagnosing and updating your installation, and the worked tutorials.

Tutorials

For guided, end-to-end walkthroughs, see the tutorials:


See also: Running and output | Parameter grids | Postprocessing and chemistry | Configuration file | Diagnose and update | Troubleshooting