Learn R Programming

mgcv (version 1.9-5)

options.mgcv: Options in mgcv

Description

Some mgcv options can be set using options. These are described here.

Arguments

Author

Simon N. Wood ([email protected]).

Details

The following options control smoother behaviour.

  • mgcv.s.bs give the default basis used for s terms. Defaults to c("tp","tp"), the first being the default for univariate smoothing and the second for mutivariate.

  • mgcv.te.bs is the default basis for te and ti smooths. Defaults to "cr".

  • mgcv.te.np is the default np parameter for te and ti. Defaults to TRUE.

  • mgcv.xt is the default xt parameter for s, te and ti. NULL by default.

  • mgcv.ti.mc is the default mc parameter for ti. NULL by default.

The following options control gam and bam methods.

  • mgcv.gam.method gives the default method argument for gam. "GCV.Cp" by default (historical, "REML" usually better).

  • mgcv.gam.optimizer gives the default optimizer argument for gam. c("outer","newton") by default.

  • mgcv.bam.discrete gives the default discrete argument for bam. FALSE by default (historical, TRUE usually better).

The following controls plotting defaults.

  • mgcv.plot.scheme controls the plotting scheme argument to plot.gam. Defaults to 0.