Produces weights for aggregates based MIDAS regression
amweights(p, d, m, weight = nealmon, type = c("A", "B", "C"))a vector of weights
parameters for weight functions, see details.
number of high frequency lags
the frequency
the weight function
type of structure, a string, one of A, B or C.
Virmantas Kvedaras, Vaidotas Zemlys
Suppose a weight function \(w(\beta,\theta)\) satisfies the following equation: $$w(\beta,\theta)=\beta g(\theta)$$
The following combinations are defined, corresponding to structure types
A, B and C respectively:
$$(w(\beta_1,\theta_1),...,w(\beta_k,\theta_k))$$
$$(w(\beta_1,\theta),...,w(\beta_k,\theta))$$
$$\beta(w(1,\theta),...,w(1,\theta)),$$
where \(k\) is the number of low frequency lags, i.e. \(d/m\). If the latter value is not whole number, the error is produced.
The starting values \(p\) should be supplied then as follows: $$(\beta_1,\theta_1,...,\beta_k,\theta_k)$$ $$(\beta_1,...,\beta_k,\theta)$$ $$(\beta,\theta)$$