Fast bounds on discrepancy

Description

Code from [Thi01]. This tool approximates upper and lower boundaries on the Linf discrepancy. As those approximations can be computed faster than the true value, it allows to approximate the Linf discrepancy of higher dimensional point sets. Can be used in higher dimensions.

Files

src/metrics/BoundariesStar.cpp  
include/utk/metrics/BoundariesStarDiscrepancy.hpp

Usage

Exe C++ Python

BoundariesStarDiscrepancy calculator
Usage: ./BoundariesStar [OPTIONS]

Options:
  -h,--help                   Print this help message and exit
  -i,--input TEXT:FILE ... REQUIRED
                              Input file(s)
  -o,--output TEXT            Output file (empty is stdout)
  --silent                    Silence UTK logs

#include <utk/utils/PointsetIO.hpp>
#include <utk/utils/Pointset.hpp>
#include <utk/samplers/SamplerWhitenoise.hpp>
#include <utk/metrics/BoundariesStarDiscrepancy.hpp>

int main()
{
    utk::Pointset<double> pts;

    utk::SamplerWhitenoise wn(2 /* dimension */);
    wn.setRandomSeed(/* empty means random, can also pass a number */);
    // Check for no errors
    if (wn.generateSamples(pts, 1024 /* Number of points */))
    {
        auto rslt = utk::BoundariesStarDiscrepancy().compute(pts);
    }
}

import pyutk
import numpy as np

dmin, dmax = pyutk.BoundariesStarDiscrepancy().compute(np.random.uniform(0, 1, (128, 2)))

License