Mort1Dsmooth_optimize: Optimize a 1D Penalized-Poisson IWLS over smoothing...
In MortalitySmooth: Smoothing and Forecasting Poisson Counts with P-Splines

Description Usage Arguments Details Author(s) References See Also

This is an internal function of package MortalitySmooth which optimizes the smoothing parameter for penalized B-splines within the function Mort1Dsmooth.

Mort1Dsmooth_optimize(x, y, offset, wei,
                      psi2, B, DtD, a.init,
                      MON, TOL1, TOL2, RANGE, MAX.IT,
                      MET)

`x`	vector for the abscissa of data.
`y`	vector of counts response.
`offset`	vector with an a priori known component (optional).
`wei`	an optional vector of weights to be used in the fitting process.
`psi2`	an overdispersion parameter used the quasi-likelihood approach.
`B`	B-splines basis.
`DtD`	inner product of the difference matrix
`a.init`	initial coefficients
`MON`	Logical switch indicating if monitoring is required.
`TOL1`	The tolerance level in the IWLS algorithm.
`TOL2`	difference between two adjacent smoothing parameters in the (pseudo) grid search, log-scale.
`RANGE`	range in which smoothing parameter should be searched.
`MAX.IT`	the maximum number of iterations
`MET`	the method for controlling the amount of smoothing

The function aims to find the optimal smoothing parameter within the given RANGE in Mort1Dsmooth with method equal to 1 or 2 (BIC and AIC). It employs the function cleversearch from package svcm in two separate steps. First it searches using a rough grid (4 times TOL2) and the median of RANGE as starting lambda. Afterwards it searches in the restricted range around the sub-optimal smoothing parameter, using a finer grid defined by TOL2.

This procedure allows to find a precise smoothing parameter in an efficient way: we do not explore the full range of possible lambda values, moving at most one grid step up or down. Furthermore the two steps routine reduces the risk of finding sub-optimal smoothing parameter.

Carlo G Camarda

Camarda, C. G. (2012). MortalitySmooth: An R Package for Smoothing Poisson Counts with P-Splines. Journal of Statistical Software. 50, 1-24. http://www.jstatsoft.org/v50/i01/.

Mort1Dsmooth_update, Mort1Dsmooth_estimate, Mort1Dsmooth.

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