Create Individual dataset for simulation

[Ravua1992]

Hi,

I am typically create my dataset using the following code, where I sample individual PK estimates from the pkpd dataset (data_pkpd):

NSIM=100
idata_SIM0<- data.frame(expand.idata(ID=c(1:nsim)) %>%
                          mutate(GRP=1) %>%
                          # mutate(WT=rep(each=200,seq(41,120,1))) %>%
                          mutate_random(CL ~ sample(data_pkpd$ICL, size=nsim, replace = TRUE))%>%
                          mutate_random(Q ~ sample(data_pkpd$IQ, size=nsim, replace = TRUE))%>%
                          mutate_random(V2 ~ sample(data_pkpd$IV2, size=nsim, replace = TRUE))%>%
                          mutate_random(V3 ~ sample(data_pkpd$IV3, size=nsim, replace = TRUE))%>%
                          mutate_random(V7 ~ sample(data_pkpd$IV7, size=nsim, replace = TRUE))%>%
                          mutate_random(Q2 ~ sample(data_pkpd$IQ2, size=nsim, replace = TRUE))%>%
                          mutate_random(KA ~ sample(data_pkpd$IKA, size=nsim, replace = TRUE))%>%
                          mutate_random(F1 ~ sample(data_pkpd$IF1, size=nsim, replace = TRUE))%>%
                          mutate_random(BL_PD ~ sample(data_pkpd$BL_PD, size=nsim, replace = TRUE))%>%
                          mutate(time=0)) 

However, mutate_random is a part of dmutate package and I am unable to use it for the current analysis.
Is there any alternative approach to create individual dataset using the mrgsovle package ?

Thank you!

[kylebaron]

You could just resample rows from your source data set

new_data <- slice_sample(data_pkpd, n = nsim, replace = TRUE) %>% mutate(ID = row_number())

I think the code using mutate_random is independently resampling your parameters? If you wanted that, can you

new_data <- pivot_longer(data_pkpd, cols = c(CL:BL_PD)) %>% group_by(name) %>% slice_sample(n = nsim) %>% ungroup() ...