pclm.general: General PCLM computations

In MaciejDanko/pclmpash:

Description

Main procedure to calculate PCLM with automated step.

Usage

pclm.general(x, y, count.type = c("DX", "LX"), out.step = "auto",
  exposures = NULL, control = list())

Arguments

x	Vector with start of the interval for age/time classes.
y	Vector with counts, e.g. ndx. It must have the same length as x.
count.type	Type of the data, deaths("DX")(default) or exposures("LX".)
out.step	Age interval length in output aggregated life-table. If set to "auto" then the parameter is automatically set to the length of the shortest age/time interval of x.
exposures	Optional exposures to calculate smooth mortality rates. A vector of the same length as x and y. See reference [1] for further details.
control	List with additional parameters. See pclm.control.

Details

The function has four major steps:

1. Calculate interval multiple (pclm.interval.multiple to remove fractional parts from x vector. The removal of fractional parts is necessary to build composition matrix.

2. Calculate composition matrix using pclm.compmat.

3. Fit PCLM model using pclm.opt.

4. Calculate aggregated (grouped) life-table using pclm.aggregate.

More details for PCLM algorithm can be found in reference [1], but see also pclm.fit and pclm.compmat.

Value

The output is of "pclm" class with the components:

grouped	Life-table based on aggregated PCLM fit and defined by out.step.
raw	Life-table based on original (raw) PCLM fit.
fit	PCLM fit used to construct life-tables.
m	Interval multiple, see pclm.interval.multiple, pclm.compmat.
x.div	Value of x.div, see pclm.control.
out.step	Interval length of aggregated life-table, see pclm.control.
control	Used control parameters, see pclm.control.
warn.list	List with warnings.

Author(s)

Maciej J. Danko <[e-mail]> <[e-mail]>

References

1. Rizzi S, Gampe J, Eilers PHC. Efficient estimation of smooth distributions from coarsely grouped data. Am J Epidemiol. 2015;182:138?47.

2. Rizzi S, Thinggaard M, Engholm G, et al. Comparison of non-parametric methods for ungrouping coarsely aggregated data. BMC Medical Research Methodology. 2016;16:59. doi:10.1186/s12874-016-0157-8.

See Also

pclm.fit, pclm.compmat, pclm.interval.multiple, and pclm.nclasses.

Examples

# The examples with use of the \code{pash} object are presented in \link{pclm.fit}.
# Explicit examples of use \code{pclm.general} (especially how to use exposures) are to be written in a next package release.

MaciejDanko/pclmpash documentation built on May 14, 2019, 7:41 a.m.