Conditional pdf's and cdf's of MixAR models

Gives conditional probability densities and distribution functions of mixture autoregressive models.

Methods

mix_pdf gives a probability density, mix_cdf a distribution function. If argument x is supplied, the functions are evaluated for the specified values of x, otherwise function objects are returned and can be used for further computations, eg for graphs.

mix_pdf and mix_cdf have methods with the following signatures.

signature(model = "MixARGaussian", x = "missing", index = "missing", xcond = "numeric")

Return (as a function of one argument) the conditional density (respectively cdf), \(f(x|xcond)\), of \(X_{t+1}\) given the past values xcond. The values in xcond are in natural time order, e.g. the last value in xcond is \(x_{t}\). xcond must contain enough values for the computation of the conditional density (cdf) but if more are given, only the necessary ones are used.

signature(model = "MixARGaussian", x = "numeric", index = "missing", xcond = "numeric")

Compute the conditional density (respectively cdf) at the values given by x.

signature(model = "MixARGaussian", x = "numeric", index = "numeric", xcond = "missing")

Compute conditional densities (respectively cdf) for times specified in index. For each \(t\in{}\)index the past values needed for the computation of the pdf (cdf) are ...,x[t-2],x[t-1].

signature(model = "MixARgen", x = "missing", index = "missing", xcond = "numeric")

signature(model = "MixARgen", x = "numeric", index = "missing", xcond = "numeric")

signature(model = "MixARgen", x = "numeric", index = "numeric", xcond = "missing")

Author

Georgi N. Boshnakov

See also

mix_moment for examples and computation of summary statistics of the predictive distributions

mix_qf for computation of quantiles.