Precision of multinomial coefficients

[Enchufa2]

There could be precision issues for high-order terms. At some point we need to check this calculation and try better approaches. Some ideas here: https://stackoverflow.com/questions/39799303/binomial-coefficients-rounding-error. We could check also if there are any analytical approximations.