partab.character: Create a Parameter Table from Model Name

Description

Creates a parameter table from a model name. Pass the project argument or set the project option.

Usage

# S3 method for character
partab(
  x,
  verbose = FALSE,
  lo = "5",
  hi = "95",
  metafile = modelpath(x, "def", ...),
  xmlfile = modelpath(x, "xml", ...),
  ctlfile = modelfile(x, ...),
  bootcsv,
  strip.namespace = TRUE,
  skip = 28,
  check.names = FALSE,
  digits = 3,
  ci = TRUE,
  open = "(",
  close = ")",
  sep = ", ",
  format = TRUE,
  fields = getOption("fields", default = c("symbol", "label", "unit")),
  relative = TRUE,
  percent = relative,
  nonzero = TRUE,
  shrinkage = FALSE,
  correlation = FALSE,
  ...
)

Value

object of class partab, data.frame

Arguments

x: a model name (numeric or character)
verbose: set FALSE to suppress messages
lo: the PsN bootstrap lower confidence limit (%)
hi: the PsN bootstrap upper confidence limit (%)
metafile: optional metadata for parameter table (see also: fields)
xmlfile: path to xml file
ctlfile: path to control stream
bootcsv: path to PsN bootstrap_results.csv
strip.namespace: whether to strip e.g. nm: from xml elements for easier xpath syntax
skip: number of lines to skip in bootstrap_results.csv
check.names: passed to bootstrap reader
digits: limits numerics to significant digits (use NULL to suppress)
ci: combine bootstrap lo and hi into an enclosed interval
open: first character for bootstrap interval
close: last character for bootstrap interval
sep: separator for bootstrap interval
format: format numerics as character
fields: metadata fields to read from control stream. See details.
relative: transform standard errors to relative standard errors: rse replaces se
percent: if relative is true, express as percent (else ignore): prse replaces se
nonzero: limit random effects to those with nonzero estimates
shrinkage: whether to include percent shrinkage on random effects
correlation: whether to include correlation of random effects (as percent if percent is true)
...: passed to other functions

Details

Normally you can just call the generic. Suitable defaults are supplied, but much customization is supported by means of arguments documented here and in called functions.

Metadata can be added to the parameter table two ways: as markup in the control stream, and as a *.def file in the model directory. See vignette('parameter-table') for details.

Examples

Run this code

library(magrittr)
options(project = system.file('project/model',package='nonmemica'))
1001 %>% partab
1001 %>% partab(shrinkage = TRUE, correlation = TRUE)