Welcome to the GitHub page for kmc_dn (Kinetic Monte Carlo for Dopant Networks). This page houses a simulation tool that was specifically built to simulate Silicon Dopant networks. In practice, this tool simulates variable range hopping between an arbitrary number of acceptor sites in a domain surrounded by an arbitrary number of electrodes. It is mostly written in Python, with (soon) some of the more computationally heavy parts written in Go.
A good introduction to the code and the original experimental context can be found in the main author's Master thesis (WHERE?).
To get started with the code, please have a look at the scripts in the examples folder. They are well documented examples that should tell you enough to get going.
Installing the tool boils mainly down to meeting a few dependencies. A suggested workflow (based on conda) is given afterwards.
- python 3.6+
- numpy
- matplotlib
- numba
- ffmpeg (for animations)
- fenics, installation notes are here
- logging
- go
The recommended way to manage the dependencies is through a virtual environment in conda. The code can run perfectly without conda, but it is a neat way to isolate the dependencies and to make sure the code will always run. For the sake of example, let's create an environment named kmc. Run the following command (in the terminal) to create the environment:
conda create -n kmc
Now activate the environment:
conda activate kmc
If this doesn't work
source activate kmc
And run the following installs:
conda install numpy
conda install matplotlib
conda install numba
conda install -c conda-forge fenics
conda install -c anaconda seaborn
Lastly, to make sure that kmc_dopant_networks is always found when running code in the kmc environment, look for the system path that looks like this:
~/anaconda3/envs/kmc/lib/python3.6/site-packages/
Then make a new file kmc.pth
which contains the absolute path to the
repo.
To use go functionalities you have to compile and build a library accessible by python inside the goSimulation folder.
go build -o libSimulation.so -buildmode=c-shared simulationWrapper.go simulation.go probabilitySimulation.go
You are now all set to start simulating! If documentation in the code and the example files are not enough, please do not hesitate to contact any of the authors below.
Bram de Wilde / brambozz; b.dewilde-1@student.utwente.nl