This is a minor PR to address a small issue unearthed whilst writing up the derivations of CSFS. It transpires that a duplication matrix and elimination matrix were the wrong way around in the derivative of vech(D_k) with respect to vech(\Lambda_k). As these matrices are already extremely similar this fix made little practical difference.
This is a minor PR to address a small issue unearthed whilst writing up the derivations of CSFS. It transpires that a duplication matrix and elimination matrix were the wrong way around in the derivative of vech(D_k) with respect to vech(\Lambda_k). As these matrices are already extremely similar this fix made little practical difference.