search for: shfn

...p[1]/(1+exp((p[2]-times)/p[3])) elliptic(data,model=mu,preg=c(17,50,8)) # variance increases with power of absolute fitted values: model fm1 # note: function for log(variance) is fitted shape <- function(p,mu) p[1]*log(abs(mu))+p[2] elliptic(data,model=mu,preg=c(17,50,8),varfn=shape,pvar=c(1,2),shfn=T) # CAR(1): model fm2 elliptic(data,model=~Asym/(1+exp((xmid-times)/scal)),preg=c(17,50,8),par=0.5) # fit the discrete time AR(1) #elliptic(data,model=~Asym/(1+exp((xmid-ctimes)/scal)),preg=c(17,50,8),par=0.5) # CAR(1) and variance depending on mean elliptic(data,model=mu,preg=c(19,55,8),varfn=sh...