This package uses Markov Chain Monte Carlo to calculate the most probable graphs for a given number of nodes in space.
Markov chain Monte Carlo (MCMC) methods are a class of algorithms for sampling from a probability distribution based on constructing a Markov chain that has the desired distribution as its equilibrium distribution. The state of the chain after a number of steps is then used as a sample of the desired distribution. The quality of the sample improves as a function of the number of steps.
You will first need to clone the repo.
git clone https://github.com/hgandhi2411/MCMC-Graphs
Then inside the repo, install the package. The usage is described in the next section.
$ npm install --save mcmc-graphsYou can use the package in your code using the following line:
const mcmcGraphs = require('mcmc-graphs');Once the repo is cloned and package is installed, run the code:
$ node lib\index.js You can use the commandline options to give the following parameters:
- -n, --nodeNumber : Number of nodes, default value is 10.
- -r, --rfactor: The r multiplier in the theta function, default is 1.
- -t, --temperature : Temperature, default value is 298.
- -c, --coordinates : Coordinates as an array, for example, -c [[1, 2], [2, 3], [4,5]]. The array size must equal the nodeNumber. The program will not throw an error if nodeNumber < 10 and coordinates are not specified, but any nodeNumber > 10 will throw an error.
An example of running the code is:
$ node lib\index.js -r 2 -n 5 -c [[1, 2], [3, 4], [6, 7], [7, 8], [0, 5]]Before running the code, uncomment the last two lines in lib\index.js before module.export.
// var top10 = mcmc();
// console.log(top10);Do not uncomment these while running the tests.
- The code throws errors after a certain number of iterations. Error still untracked!
- Quantiling not as expected. The package is still broken, it doesn't give the top 10% of graphs correctly.
- Tests for mcmc function not in place
Apache-2.0 © Heta Gandhi