A small implementation of Random Boolean Networks evolving in time.
After having seen this video, I felt inspired and I decided to implement a small, similar project to understand better this model and also as an excuse to practice with Golang 😉
With respect to the video, I decided to change the degree of randomness: instead of looking up the new value on a truth table, I decided to flip the value of the node (from 0 to 1 or viceversa) based on the list of neighbors. As a direct consequence, during an evolution step a node can change its value multiple times.
There are 3 values to play with in your experiments:
EPOCHS: how many time units the model evolves
LINKS : how many neighbors each node has in the DiGraph
NODES : how many nodes in the DiGraph
Notice that, in order to produce a squared image, NODES should be a perfect square and also a power of 2, while links whould be the exponent of the latter. To summarize in a simple equation:
First of all you have to clone this repository:
git clone https://github.com/A-725-K/GoRaBoN.git
Than you have to set your desired parameters in file main.go and launch it with the following command:
go run main.go [-epochs n] [-links n] [-nodes n]
You can specify on the command line 3 options:
-
-epochs <num>: number of iterations, must be
$\geq 1$ -
-links <num>: number of neighbors of each node in the network, must be
$\geq 1$ - -nodes <num>: number of nodes in the Random Boolean Network, must follow the previous equation [1].
Finally, to get an animated gif like the one at the beginning of this README, you have to install the requirements and launch the Python script in the directory plot:
cd plot
pip install -r requirements.txt
python3 plot_results.py <EPOCHS>
where <EPOCHS> must match the parameter given to the Go script.
Enjoy yourself! 😄
- Andrea Canepa - Computer Science, UNIGE