Kalman filter
[betat,betaf,sigmatt,sigmatf]=filter(param,func,y,x,F,z)
* param = a vector of parameters (sqrt of variances)
* func = the function which transforms the parameters into the matrix of variances (Q and R) and, if necessary, into the matrices of prior values
* y = (nx1) data vector
* x = (nxk) data matrix
* F = transition matrix
* z = (nxl) data matrix
* betat = vector of filtered beta at date t with the information available at date t (beta(t|t))
* betaf = vector of filtered beta at date t with the information available at date t-1 (beta(t|t-1)
* ferror = vector of errors at date t ( = y-x*beta(t|t))
* sigmatt = vector of filtered variances at date t with the information available at date t (sigma(t|t))
* sigmatf = vector of filtered variances at date t with the information available at date t-1 (sigma(t|t-1))