I wish to fit a dynamical model in R and I am running in a problem that requires some of your wisdom to solve. For SAS users I am searching for the equivalent of the retain statement. For people that want to read complicated explanations to help me: I have a system of two equations written as difference equations here. To boil it down. I have a dataframe with three variables y, X1, X2 which are measured. Now, I want to estimate a model which says y= yhat+epsillon as in standard nonlinear regression. Where yhat is the predicted value of y. yhat can be calculated as follows: # This code illustrates how I could simulate the expected values of y if I knew the values of the parameters tau and b. # but in reality I would like to estimate them. # code is for illustration of the principles and is not meant to be functional!! yaux[1]<-0 b<- a_number # b would have to become estimated by nls or nlme tau<- another_number # tau would also be estimated in nls or nlme for (t in 2:1000) { yaux[t+1] <- yaux[t] + (X1-yaux[t])/tau yhat[t+1] <- yaux[t+1]*X2[t+1]/(X2[t+1]+b) } Now, my problem is that I do not know the values of /tau/ and /b/ and I would like to estimated them by non-linear regression. This is easy in the case of /b/ but for /tau /nls (or nlme etc) would have to remember the value of /yaux /for the previous observation and I did not find any syntactical mean to do that. [[alternative HTML version deleted]]