A general puprose tool for modelling reaction-diffusion systems in MATLAB (with extensible support for other computational environments). Inspired by the very fun and useful dynamical systems analysis tool xppaut, we have aimed to make chemical kinetic systems more approachable by providing a tool that automatically derives sets of equations from reactions schemes and can perform some of the more tedious aspects of model simplification in an reliable and automated manner.
Moreover, due to our reaction-based model specification format, combining models is often as simple as copy-pasting the contents of two files.
A longer term aim of this project is to provide a tool that separates model specification from model analysis, simulation building, and simulation analysis.
Currently, we have developped a simple text-based model specification paradigm for reaction-diffusion systems that is flxeible to be extended to consider other types of systems (e.g., reaction-diffusion-convection).
N.B. All code is written from MATLAB 2018+ (i.e., array braodcasting is used), but one may be able to run it on older versions with some elbow grease.
This library provides three main components that are useful for CPM simulations:
- An extensible model specification format and a parser that will generate model-specific files for running CPM simulations.
- A core set of functions/scripts for running CPM simulations.
- A basic system for switching between models, running simulations, and storing results in a systematic manner.
The repository should be downloaded and placed somewhere that MATLAB can access (and added to its path).
One may use the following commands in the command line to download and open the library:
git clone https://github.com/laurentmackay/CPM_RD_matlab.git
cd CPM_RD_matlab
matlab
Do not forget to add the base folder to your MATLAB path!!!!
As a basic example, try the entering the following commands into the MATLAB command line:
deploy_model('chem_Rx_Pax_Asheesh')
main_FVM
Eventually, a window should pop up with CPM cell that moves around. Check the _chem_Rx_Pax_Asheesh folder for .ode and .f90 files that can be used with xppaut and AUTO, respectively.
You may also have a look at the script main_FVM.m, to see if there are any simulation parameters you wish to change.
After creating a model specification file (e.g., in the file models/model_file) for your model, one should:
- Deploy the model
- Create a "virtual experiment" to help organize results on the file system using
set_experiment(). [Optional] - Run a CPM simulation using
main_FVM.m - Analyze the results using tools in /protocols/CPM/analysis/
A model specified in the file models/model_file can be used in simulations by first calling deploy_model('model_file'). This will parse the model specified in models/model_file and generate model-dependent functions necessary for running simulations. These generated files are stored in a directory called _model_file and that directory is temporarily added to the MATLAB-path.
Only one model can be deployed at a time, and deploy_model will take care of modifying the MATLAB-path when a user switches between models. One may check the currently active model using the active_model global variable.
Moreover, deploy_model will also check if the model specification file has been updated since it was last deployed, and will only run the model through the interpreter when changes have been made. This behaviour can be overridden by passing a second boolean argument true or 1 to deploy_model.
In order to prevent overwriting of files and disroganized data, we use "experiments" to define where results should be saved. In order to set the name of your current experiment, use the command set_experiment('experiment_name').
The current results directory can be obtained using results_dir(), and, in general, is given by <CPM_RD_matlab>/_model_name/results/experiment_name where <CPM_RD_matlab> is the path to the base directory of this library.
The function ls_results() can be used to check the results directory for .mat files, or one may also use ls_results(ext) to find all files that end in .ext.