Select an LLLM at each point

Share:

Description

Select LLLMs for each row of the input data.

Usage

1
apply.ic.fit(ydens, models, ess, mct, ic, cell.adj, averaging, loud = TRUE)

Arguments

ydens

A matrix with 2^k-1 columns, one for each capture pattern. Each row sums to 1; these are empirical capture pattern probabilities.

models

A list of character vectors, with each vector containing column names from the associated log-linear design matrix. For example, see the output of make.hierarchical.term.sets().

ess

A vector of effective sample sizes, one for each row of ydens.

mct

The number of population units that were observed for each row of ydens.

ic

The chosen information criterion. Currently implemented: "AIC", "AICc", "BIC", "BICpi".

cell.adj

Logical: TRUE means that the cell adjustment of Evans and Bonet (1995) is applied.

averaging

Logical: TRUE means that the information criterion weights are used to do model averaging, locally.

loud

Logical: TRUE means that the progress is noted by printing the number of the row of ydens currently being processed.

Details

See Kurtz (2013). Each row of ydens corresponds to a covariate vector, and contains a local average of multinomial capture pattern outcomes across nearby points. apply.ic.fit applies the function ic.fit at each row. The vector of local effective sample sizes is crucial, and is specified in the ess argument.

Value

lll

An object of class "lllcrc"

Author(s)

Zach Kurtz

References

Kurtz (2013)

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.