crossValFit: Cross Validation for Model Fits

In hhabra/Combiner: Method for Combining LC-MS Metabolomics Feature Measurements

Cross Validation for Model Fits

Description

Helper function for fit_gam() & fit_loess(). Determines optimal value of k basis functions for Generalized Additive Model fits or span for loess fits from among user-defined choices, using a 10-fold cross validation minimizing mean squared error.

Usage

crossValFit(
  rts,
  fit,
  vals,
  bs,
  family,
  m,
  method,
  optimizer,
  control,
  message,
  ...
)

Arguments

rts	data.frame of ordered pair retention times
fit	Either "gam" for GAM fits, or "loess" for loess fits
vals	numeric vector: k values for GAM fits, spans for loess fits. Best value chosen by 10-fold cross validation.
bs	character. Choice of spline method, either "bs" or "ps"
family	character. Choice of mgcv family; see: ?mgcv::family.mgcv
m	integer. Basis and penalty order for GAM; see ?mgcv::s
method	character. Smoothing parameter estimation method; see: ?mgcv::gam
optimizer	character. Method to optimize smoothing parameter; see: ?mgcv::gam
control	control parameters for loess fits; see: ?loess.control
message	Option to print message indicating function progress
...	Other arguments passed to mgcv::gam.

Value

Optimal parameter value as determined by 10-fold cross validation

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