Stata 15 help for mf_ucmpsdensity


[M-5] ucmpsdensity() -- Parametric spectral density of a time-series stochastic cycle


real matrix ucmpsdensity(real scalar n, real vector rho, real vector lambda, real vector order, real scalar v, real scalar pspectrum, |real vector range)


ucmpsdensity(n, rho, lambda, d, v, pspectrum, |range) computes the parametric spectral density of a time-series stochastic cycle defined by its order, frequency, and damping effect. The cycle frequencies are stored in the vector rho, and the damping effects are stored in the vector lambda. v contains the process idiosyncratic error variance. These cycle parameters can be extracted from an unobserved-components model.


ucmpsdensity() returns an n x k+1 matrix containing the spectral densities in the first k columns and the frequencies in the last column, where k = length(rho). The parametric spectrum is computed if pspectrum != 0. By default, range = (0,pi())


ucmpsdensity(n, rho, lambda, order, v, pspectrum, |range): n: 1 x 1 rho: k x 1 or 1 x k lambda: k x 1 or 1 x k order: k x 1 or 1 x k v: 1 x 1 pspectrum: 1 x 1 range: 2 x 1 or 1 x 2 result: n x k+1


The frequencies, rho, must be in (0,pi()). The damping effects, lambda, must be in (0,1). The cycle order, order, must contain positive integers. The variance parameter, v, must be greater than zero. The range, if specified, must be in [0,pi()].

Source code


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index