Mort2Dsmooth_se: Compute a 2D standard errors
In MortalitySmooth: Smoothing and Forecasting Poisson Counts with P-Splines

Description Usage Arguments Details Value Author(s) See Also

This is an internal function of package MortalitySmooth which calculates the inner product of a matrix (from a Kronecker product) and a sparse weight matrix in order to obtain standard errors. It uses the same idea employed in MortSmooth.BWB and the elements after the IWLS converged, including the penalty term.

1	Mort2Dsmooth_se(RTBx, RTBy, nbx, nby, BWB.P1)

`RTBx`	tensors product of B-splines basis for the x-axis.
`RTBy`	tensors product of B-splines basis for the y-axis.
`nbx`	number of B-splines for the x-axis.
`nby`	number of B-splines for the y-axis.
`BWB.P1`	inverse of the LHS of the Poisson system of equations.

This function is only used within predict.Mort2Dsmooth when standard errors are required. The arguments BWB.P1 is the LHS after convergence is reached and smoothing parameter selected. The standard errors as given in the function are computed for the linear predictor term and simple computation is needed to obtain standard errors for the Poisson counts. Anyway predict.Mort2Dsmooth takes care of such differences.

The Generalized Linear Array Models setting is explained in the reference in MortSmooth_BWB and Mort2Dsmooth.

A matrix of standard errors for the linear predictor term.

Carlo G Camarda

Mort2Dsmooth, MortSmooth_BWB, predict.Mort2Dsmooth.

MortalitySmooth documentation built on May 2, 2019, 6:07 a.m.