Bayesian Blocks

Header-only C++ implementation of the Bayesian Blocks argorithm. Plug and play!

#include "baeysian_blocks.hpp"
#include "baeysian_blocks_root.hpp"  // if you want to rebin ROOT CERN histograms

int main() {

    // "data" is a a std::vector containing the observed values
    // [optional] "weights" can be used to specify the weight (how many times an observation occurs)
    auto change_points = BayesianBlocks::blocks(data, /* optional */ weights);
    // "change_points" is a std::vector
 
    // "hist" is a ROOT TH1*
    auto h_rebin = dynamic_cast<TH1*>(BayesianBlocks::rebin(hist));
    // "h_rebin" is the rebinned histogram
    
    return 0;
}

Check out the BayesianBlocks::blocks and BayesianBlocks::rebin signatures for a more advanced usage. Have a look at test/run_test.cc too.

Bayesian blocks algorithm reference: Scargle, J et al. (2012) [https://doi.org/10.1088/0004-637X/764/2/167]

The bblocks utility

The bblocks program rebins all the 1-dim histograms contained in a ROOT file.

Install:

$ cd cpp
$ PREFIX=<install-prefix> make install

Run:

$ bblocks --help
USAGE: bblocks [-v|--verbose] [-h|--help] [--p0 <val> (default 0.01)] file1 file2 ...

Related

Julia language enthusiast? Checl out my Julia package: https://github.com/gipert/BayesianBlocks.jl