Summarise model parameters in a table — get_par

Builds a data.frame combining initial values, estimated values, bounds, gradient, gradient check, and bounds check for all parameters. Each row is one scalar element of the unlisted parameter vector.

Usage

get_par_table(
  obj,
  parameters,
  map,
  lower = NULL,
  upper = NULL,
  grad_tol = 1e-04,
  bnds_tol = 0.025,
  include = c("core", "all_est", "all"),
  digits = 3,
  show_map = FALSE
)

Arguments

obj: An RTMB AD object from MakeADFun.
parameters: Named list of initial parameter values passed to MakeADFun.
map: Named list of map factors passed to MakeADFun.
lower: Numeric vector of lower bounds (length length(obj$par)), or NULL for -Inf everywhere.
upper: Numeric vector of upper bounds (length length(obj$par)), or NULL for Inf everywhere.
grad_tol: Threshold for gradient check: absolute gradient below this is "OK", otherwise "BAD". Default 1e-4.
bnds_tol: Fraction of finite bound range defining the proximity margin for the bounds check. Default 0.025 (2.5%).
include: Which rows to return: "core" (default) returns estimated parameters excluding rdev_y; "all_est" includes rdev_y; "all" includes fixed parameters too.
digits: Number of significant digits for numeric columns. Default 3. Use NULL to suppress rounding.
show_map: Logical. Include map and fixed columns. Default FALSE.

Value

A data.frame with columns par, init, est, lwr, upr, grd, gr_chk, bd_chk, and optionally group, map and fixed (when show_map = TRUE).