ofxBayes

Introduction

ofxBayes is openFrameworks addon for Bayesian statistics. Now this is a baby project. Currently, you can

• Add your own hypothesis and probability as map<int, int>
• Write your own likelihood function by giving lambda function.

for detail take a look at codes below.

Dependency

currently none.

Example

example-basic

Simple Monty-hall problem. Three steps required.

1. Make prior probability
2. Write likelihood function
3. Update your guess!

    //Hypothesis and data
    _map[0] = 0.f;
    _map[1] = 0.f;
    _map[2] = 0.f;

    //initialization
    ofxBayesHist hist(_map);
    hist.initialize(); 
    hist.disp("Prior probability");

    //Likelihood function
    int choise = 0;
    hist.likelihoodFunc = [=](int data, int hypo) { 

        if (hypo == data)
            return 0.f;
        else
        {
            if (hypo == choise)
                return 0.5f; 
            else
                return 1.f;
        }
    };

    //update posterior prob.
    hist.update(1);
    hist.disp("Posterior probability");

License

MIT

Acknowledgement

Lot of inspiration from http://greenteapress.com/wp/think-bayes/