Given mean and covariance matrix of a multivariate normal distribution plus a part of observations, I really need to generate the conditional normal given these distributions. The most ideal package I have found is r-condMVNorm, which contains the method rcmvnorm (see for example https://rdrr.io/cran/condMVNorm/man/cmvnorm.html) doing exactly what I want.

However, it has come the embarrassing fact that this package is not available in conda forge, whereas due to programmatic reasons I seriously prefer this to be done on a package that can be found on conda forge (or, if possible, r-base or r-MASS). Is there a solution or will I have to implement the conditional normal on my own ?