sf: Defining smooth functions in semiparametric model formulae
In vows: Voxelwise Semiparametrics

This function is called by semipar.mp to define B-spline smooths.

1 2	sf(argvals, effect = NULL, k = 10, norder = 4, pen.order = 2, range.basis = NULL, knots = "quantile")

`argvals`	a vector or matrix of covariates.
`effect`	predictor whose effect varies with respect to `argvals`. E.g., if the effect of `diagnosis` varies with `age`, use `sf(age, effect = diagnosis)`. Similar to argument `by` in `s`.
`k`	number of B-spline basis functions.
`norder`	order of B-splines: the default, `4`, gives cubic B-splines.
`pen.order`	order of the penalty, i.e., of the derivative defining the penalty.
`range.basis`	a numeric vector of length 2 defining the interval over which the B-spline basis is created. If `NULL`, set to the range of the variable.
`knots`	knots placement method for B-spline smoothing. The default, "quantile", places the knots at equally spaced quantiles of the data; "equispaced" gives equally spaced knots.