Skip to content

Using limit cycle continuation methods to speed up evaluation of ion channel cell models

Notifications You must be signed in to change notification settings

CardiacModelling/ContinuationForOptimisation

Repository files navigation

ContinuationForOptimisation

Using limit cycle continuation methods to speed up evaluation of ion channel cell models

Using the code

The code can be cloned from GitHub (most recent version), or downloaded from Zenodo (version of record).

Setting up the Julia project

For up to date versions of the dependent packages, run the packages.jl file.

To use the same package versions as used in the paper, refer to the Project.toml and Manifest.toml and use Pkg.instantiate().

Note the figures were generated under a different environment to the one given in the Project.toml and Manifest.toml files, which were only used to generated the results/data.

Running the analysis

The analysis from the paper can be run by running the scripts simulationTimings.jl, mcmcSetup.jl and mhmcmc.jl.

Note that running the scripts will override the previously recorded results.

When running mhmcmc.jl, there are a handful of changes you may wish to make. The number of samples to run the MCMC algorithm for can be changed from the default of 40,000. The "converger" can be chosen by setting the variables use_continuation and use_tracking_ODE. If use_continuation=true, then the continuation converger will be used. Otherwise, the ODE solver will be used to converge, running from a previous limit cycle if use_tracking_ODE=true or from standard initial conditions otherwise.

Reproducing the figures from the pre-generated data

The figures can be regenerated using figures.jl (for results figures) and diagrams.jl (for non-results figures such as the phase plane diagrams).

About

Using limit cycle continuation methods to speed up evaluation of ion channel cell models

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages