No replies to this topic

[Teodora Dumitrescu]

Hi! I would like to code the residual error for observations in a peripheral compartment so that I have an additive error when the cell number at a particular timepoint is > = 1 (*10^7) and a proportional error when the cell number is < 1 (*10^7). My code is: covariate(pbmccc1) # pbmccc1 this is the data item in my dataset that gives me the cell count IND=pbmccc1>=1 # IND should take a value of 1 when the number of cells is >=1 (*10^7) or zero when the number of cells in <1 (*10^7) observe(CObs=C + CEps * (IND + C*(1-IND)*CMixRatio)) My questions are: 1) since this is the only place where the cell count (pbmccc1) is used in the model (I am not using it as a covariate per se), is this code appropriate? 2) If not, can you suggest any other way to read into the model the information on cell count? 3) Do you see any other problems with this code? Thanks, Dora