This is what I mean. s1, s2, s3 - are my x's.
???? The s1 etc. are functions (of n) and not arguments of your functions f or F, so they are not your problem.
The problematic part may be the alpha, beta, etc. and as it was already said, the solution is to turn them into vectors. You may as well follow my first suggestion and split the expression into several subexpressions but I think in your case the vector appraoch is the better solution. You should also create a function s(n) which returns a vector consisting of the values s1(n) .. s4(n), the first elelement being 1. That way you can write your function in a very compact way by using the scalar vector product instead of the sums:
Or if you like you may even reduce the function to a single (matrix) argument ;-)
This approach assumes the gamma will have the same number of elements as the other vectors (alpha .. eta). In fact given from your equations this is not a must. The size gamma must be 1+number_of_other_vetcors. So if gamma has different dimension than alpha, etc, either the Matrix M would have to be constructed differently or (easier) you function f should be defined as function of two arguments (gamma and a matrix with all the others).
Nevertheless I still think you should create a new thread or maybe Ryan can help and create a fork here.