Bayesian Inference and Optimisation for the Monash electrochemical Simulator (MECSim) is the application developed by the monash electrochemistry group with the assistance of Assosiate Professor Jie Zhang, Emeratious Professor Alan Bond and technical assistance from Dr Gareth Kennedy And Dr Martin Robinson.

It is an automatic plaform for parameterisation that uses mathematical optimisation and Bayesian Inference to calculate parameters involved in the electrochemical simulation. Built around MECSim and first applied in the paper titled A Comparison of Bayesian Inference Strategies for Parameterisation of Large Amplitude AC Voltammetry Derived from Total Current and Fourier Transformed Versions. BIOMEC allows for automated parameterisation of DC and FTAC voltammetry, allowing highly dimensional fits of the posterior distribution.

For an in depth tutorial on installation, application and analysis of BIOMEC and its outputs watch the four part series on Youtube.

BIOMEC uses PINTS for univariant Bayesian inference.

For information of current uses see the original BIOMEC paper, the original Bayesian inference paper for AC voltammetry or our most recent featured article.

The code is run in a singularity container which works for Ubuntu/UNIX and MAC (untested) OS systems. Singularity will need to be installed to use the image. Where the guide is seen in the following website or downloaded from connected singularity hub.

Once singularity has been installed, download the BIOMEC file and run the code to create the BIOMEC container (which should be around 580MB).

$ sudo singularity build BIOMEC.simg Singularity.def

Once the image is built the input file (input.txt) can be passed to the image by using the following command.

$ ./BIOMEC.simg input.txt

or

$ singularity run BIOMEC.simg input.txt

This will generate and output file with plots and results once completed.

inputwritter.py can guide users unfamilaur with generating input files to create an input file for the BIOMEC container, this program is contained in the BIOMEC_inputwritter. Simply run the file using the following command and follow the prompts and an input file will be generated.

$ python3 inputwritter.py

The output of this file will then be of the form <input.txt> though other names will work. It is important that a copy of the MECSim Master.inp file is present in the folder you run inputwritter.py as the MECSim input file is required for BIOMEC to run.

Once comfortable with writing the input file it is recommended to use any text editor.

These are the currently supported parameters that can be treated as variables in BIOMEC for CMA-ES and Bayesian Inference calculations.

For parameters that occur on repeatable lines in the MECSim input file, change the number after the parameter code from 1 to what repeat line the parameter of interest is on.

For an in depth tutorial on installation, application and analysis of BIOMEC and its outputs watch the four part series on Youtube.

PDF tutorial may be written up later.

• BIOMEC_inputwritter: Basic terminal/ .exe code for guiding uses in writing the input files for BIOMEC
• MCMC PLOTTER: code to plot the MCMC output chains from the Iter_log.txt to images

• Custom waveforms have not been tested and Estart and End cannot equal zero.
• AC harmonic based methods require the experimental data to contain a multiple of two data points.

Please, cite the original BIOMEC paper if you have used this package in a publication.

BIOMEC analysis/python code is open source under the GPL-3.0 License, with MECSim developed by Gareth Kennedy and contaned in the mecsim.cpython-37m-x86_64-linux-gnu.so shared object is under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

For Questions/Bugs Email me at [email protected].