recursive function

Recursive estimation

Recursive estimation of coefficients and standard errors

recursive(object, spec="mean", std.errors=TRUE, from=40, tol=1e-07,
  LAPACK=FALSE, plot=TRUE, return=TRUE)

Arguments

• object: an arx, gets or isat object
• spec: 'mean' or 'variance'. If 'mean' (default), the the recursive estimates of the mean-equation are estimated
• std.errors: logical. If TRUE (default), then the coefficient standard errors are also computed
• from: integer. The starting point of the recursion
• tol: numeric. The tolerance for linear dependency among regressors
• LAPACK: logical, TRUE or FALSE (default). If true use LAPACK otherwise use LINPACK, see qr function
• plot: NULL or logical. If TRUE, then the recursive coefficient estimates are plotted. If NULL (default), then the value set by options determines whether a plot is produced or not.
• return: logical. If TRUE (default), then the recursive estimates are returned in a list

Returns

If return=TRUE, then a list is returned with the following components: - estimates: a zoo matrix with the recursive estimates

• standard.errors: a zoo matrix with the standard errors

Author(s)

Genaro Sucarrat, http://www.sucarrat.net/

See Also

ols, qr, solve.qr

Examples

##generate random variates, estimate model:
y <- rnorm(100)
mX <- matrix(rnorm(4*100), 100, 4)
mymodel <- arx(y, mc=TRUE, mxreg=mX)

##compute recursive estimates and plot them:
recursive(mymodel)