## A small point for the code [GxP / QC / QA]

Dear Helmut,

❝ for (i in 3:n)

Works for all crossover datasets, whereas

❝ for (i in subj:n) {

Works only if the sequence for the first two subjects is the same. If the first two subject's sequences are different (i.e. TR and RT), then it gives the below error, as calculations are not feasible due to zero degree of freedom.

Error in power.TOST(CV = res$CV[i - 1], theta0 = res$PE[i - 1], n = length(unique(tmp$Subject))) : n too small. Degrees of freedom <1! In addition: Warning messages: 1: In qt(a, object$df.residual) : NaNs produced 2: In anova.lm(m) :   ANOVA F-tests on an essentially perfect fit are unreliable

❝ I was only talking about importing the data from xls(x).

This is what I meant.

❝ I’m still working on the script (in the meantime a monster with 525 lines of code), allowing not only 2×2×2 but also higher-order crossovers. Not fail-safe.

Thanks in advance, looking forward to the goodie.

Regards,
Sameep.