Assume I want to predict an observable X by several other variables Y_1,...Y_k via a linear rational function (i.e., a function of the form
X ~ (c_0 + c_1*Y_1 + ... + +c_k*Y_k) / (1 + d1*X1... +c_k*Xk).
Is there some package (ideally an R package) capable of doing this? I tried it with the R function nls(), but this frequently resulted in an error or poor performance models...
Many thanks for your help in advance!
