EmpiricalDistribution#density() shouldn't return NaN if kernel density is 0.0

[axkr]

Shouldn't the "convention" be, that if kernel.density(x)equals 0.0 the EmpiricalDistribution#density() method should also return 0.0 and not NaN if kB(binIndex)==0.0 ?

hipparchus/hipparchus-stat/src/main/java/org/hipparchus/stat/fitting/EmpiricalDistribution.java

Line 581 in a8fe2cf

return kernel.density(x) * pB(binIndex) / kB(binIndex);

So the patch could be something like this?

    public double density(double x) {
        if (x < min || x > max) {
            return 0d;
        }
        final int binIndex = findBin(x);
        final RealDistribution kernel = getKernel(binStats.get(binIndex));
        double kernelDensity = kernel.density(x);
        if (kernelDensity == 0d) {
          return 0d;
        }
        return kernelDensity * pB(binIndex) / kB(binIndex);
    }

[maisonobe]

I am not a specialist of statistics, but your reasoning seems correct to me.