Bioequivalence and Bioavailability Forum

Dear all!

Thanks for your fast responses.

@Helmut: Thanks for the hint, I just contacted the authors of the package.

@ElMaestro: Thanks. Relationships are known (but complex). I tried to invert the equations by using the substitution mehtod without success. I was just surprised that the web showed no solution..

@nobody: Of course, a log-normal distribution is another possible scenario. And in fact its the same issue with function lnorm in R: you can only specify mean and sd for the underlying normal distribution. BUT: Relationships for mean and variance are much simpler between normal and log-normal distribution and therefore, solving is not that tricky:

rlnorm(n=n, meanlog=log(mu)-0.5*log(1+cv^2), sdlog=sqrt(log(1+cv^2)))

where mu and cv represent the mean and the cv for the resulting log-normal distribution (not for the underlying normal distribution).

Thanks again,
Alex