It is possible to calculate the achieved power for a given sample sizes on each arm?

[bljm]

Dear Sir:
Thank you very much for your excellent package! I was wondering if it's possible to directly obtain the achieved power given different sample sizes for each arm, while specifying a particular configuration for correction and power. For example, I tried to calculate the achieved power for the toy example you provided, as follows using the internal functions:

des <- des_ss_norm(K    = 3,
                   alpha      = 0.05,
                   beta       = 0.2,
                   delta1     = 1,
                   sigma      = rep(1, 4),
                   ratio      = rep(1, 3),
                   correction = "dunnett",
                   power      = "marginal")

comp <- build_ss_norm(n = c(16.86988,16.86988,16.86988,16.86988), alpha = 0.05, beta = 0.2,
                          delta1 = 0.5, delta0 = 0, sigma = rep(1, 3 + 1),
                          correction = "dunnett", power = "marginal",
                          summary = F)

comp$outcome<-"norm"
comp<-multiarm:::components_ss_power_setup(comp)
comp   <- multiarm:::components_ss_distribution(comp)
comp <-   multiarm:::components_ss_update(comp)

pwr_marginal <- multiarm:::root_power_ss(N=50,comp)

In the last line, I attempted to calculate the target power for a reduced total sample size of N=50. However, the root_power function doesn't seem to work as expected. Is it possible to obtain the achieved power if I specify the correct comp object, or have I perhaps misunderstood the procedure in your package?

Many thanks in advance for any clarification, and for any light on how to do this if it's possible.

Kind regards,

Johanna