'Finding the probability density function using MCMC (QUESO-library)

I've been trying to solve an inverse problem using the solveWithMetropolisHastings algorithm from QUESO library. It is a simple problem. I use the following model:

model[i][j] = e^-a[i]*t[j]

Where t = [0,0.5,1.0,1.5,2.0,2.5,3.0,3.5,4.0,4.5,5.0,5.5] and a is generated using a distribution.

The likelihood used by this problem is the following:

template<class V, class M>
double Likelihood<V, M>::lnValue(const V& domainVector) const{
    double a = domainVector[0];
    double model = exp(-a*m_t[poi]);
    double error_A = abs((model - m_data_mean.at(0))/m_stdDevs.at(0));
    double misfitValue = -1*error_A*error_A;
    return misfitValue;
}

I already tested this likelihood function and model using a as a normal and a lognormal distribution and was able to successfully solve the inverse problem. However when I generate a as an uniform distribution, the solution is a PDF of a normal distribution. I'm not sure where the issue is. If anyone has any suggestions.

c++mcmcuniform-distribution


Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source

Related Questions