fitgamParam: List fit_gam Defaults

View source: R/params.R

fitgamParamR Documentation

List fit_gam Defaults

Description

List of default parameters for GAM fitting step of main package workflow, which can be used as input for the wrapper functions. See help(fit_gam) or ?fit_gam for more details.

Usage

fitgamParam(
  useID = FALSE,
  k = seq(10, 20, 2),
  iterFilter = 2,
  outlier = "MAD",
  coef = 2,
  prop = 0.5,
  weights = 1,
  bs = "bs",
  family = "scat",
  m = c(3, 2),
  method = "REML",
  rtx = c("min", "max"),
  rty = c("min", "max"),
  optimizer = "newton",
  message = TRUE
)

Arguments

useID

choice of preserving identity-based anchors; default: FALSE

k

values for GAM basis dimension k

iterFilter

number of outlier filtering iterations; default: 2

outlier

outlier filtering method (either "MAD" (mean absolute deviation) or "boxplot"); default: "MAD"

coef

outlier filtering coefficient; default: 2

prop

minimum proportion of fits in which a point can be a flagged outlier; default: 0.5

weights

optional supplied weights to individual points; default: 1

bs

choice of spline type ("bs" or "ps"); default: "bs"

family

choice of family ("scat" or "gaussian"); default: "scat"

m

basis and penalty order; default: c(3,2)

method

smoothing parameter estimation method; default: "REML"

rtx

ordered pair of endpoints for rtx; default: ("min", "max")

rty

ordered pair of endpoints for rty; default: ("min", "max")

optimizer

numerical optimization for GAM; default: "newton"

message

option to print progress message; default: TRUE

Value

list of fit_gam parameters

See Also

fit_gam, metabCombine

Examples

fitParam <- fitgamParam(k = c(12,14,18,20), iterFilter = 1, bs = "ps",
                        family = "gaussian", method = "GCV.Cp")


hhabra/Combiner documentation built on Jan. 26, 2024, 10:30 p.m.