Problem When a correlation or covariance matrix is not positive definite (i.e., in instances when some or all eigenvalues are negative), a cholesky decomposition cannot be performed. Sometimes, these eigenvalues are very small negative numbers and occur due to rounding or due to noise in the data. In simulation studies a known/given correlation has to be imposed on an input...