mcasopt - Optimisation package for MCAS

The mcasopt Python package uses numerical optimisation routines to find approximately-optimal parameter values for MCAS, given a desired set of output resonances.

It is distributed under the terms of the MIT License (see LICENSE).

Note

This documentation should be the complete reference to using and modifying mcasopt, but it is currently a work in progress. If you want to contribute to the documentation, or report problems (such as missing or incomplete content), please file an issue on the issue tracker.

Getting started

• Installation
  • Requirements
  • Create a Python virtual environment
  • Install from source
  • Check that mcasopt is installed
  • Update from source
• Quick-start

Introduction to MCAS

• Multi-channel algebraic scattering (MCAS)
  • Selected bibliography
• Running MCAS for the first time
• Model parameters
  • Example parameters file
• Model output
• Parameter optimisation with mcasopt
  • Optimisation methods
• The error metric (objective function)
  • Functional form
  • Limitations of experimental data
  • Selecting appropriate states for comparison

High Performance Computing

• Spartan HPC system
• Slurm job scheduling system
• Gadi HPC system
  • Connecting to Gadi
  • Loading modules
  • Running jobs
  • Python virtual environments

The mcasopt package

• Overview of mcasopt
  • How to use mcasopt
  • What mcasopt does
  • What the solver does
• Running mcasopt on Spartan
  • Initial setup
  • Create an experiment directory
  • Create the required experiment files
  • Start the solver
  • Monitor the solver
  • Successful completion
  • Running from multiple starting points
  • Running your own experiments
• API documentation
  • Running the optimiser
  • Function evaluation
  • Running commands
  • Configuration settings
  • Model criteria
  • Model parameters
  • Caching simulation results
  • Example experiments
• Testing mcasopt

Contributing to mcasopt

• Contributing to mcasopt
  • Licensing
  • Workflow
• Writing documentation
  • Authoring
  • Building
• License

Indices and tables

• Index

• Module Index

• Search Page