prepare.numeric {hero}

Prepare data vector for sandwich smooth

Description

prepare.vector prepares a data vector for the sandwich smooth. Unlike the other prepare.* functions, x and splines do not need to be lists since the data are 1-dimensional.

Usage

## S3 method for class 'numeric'
prepare(data, x, splines, m = 2, sparse = TRUE, ...)

Arguments

data

A numeric data vector

x

A sequence of equidistant values corresponding to where the data are observed. Equidistant spacing between 0 and 1 is assumed if not supplied. See Details.

splines

A spline-related object, e.g., produced by bspline. A spline is automatically created if not supplied. See Details.

m

A positive integer indicating order of the difference penalty.

sparse

A logical value indicating if the result should be a sparse version of the Matrix-class.

...

Not currently implemented.

Details

If x is not supplied and n is the length(data), then the function automatically sets x = seq(0, 1, length = n).

If splines is not supplied, and n is the length(data), then the function automatically sets splines = bspline(range(x), nknots = min(ceiling(n/4), 35)).

Value

A prepared_numeric object.

Author(s)

Joshua French. Based off code by Luo Xiao (see References).

References

Xiao, L. , Li, Y. and Ruppert, D. (2013), Fast bivariate P-splines: the sandwich smoother. J. R. Stat. Soc. B, 75: 577-599. <doi:10.1111/rssb.12007>

Ruppert, D., Wand, M. P., & Carroll, R. J. (2003). Semiparametric Regression. Cambridge University Press. <doi:10.1017/CBO9780511755453>

See Also

bspline, default.evalargs, default.splines

Examples

# create data
n = 160
x = seq(0, 4 * pi, len = n)
# "true" data
mu = sin(x)
# plot true data
plot(x, mu, type = "l")
# construct noisy data
set.seed(4)
data = mu + rnorm(n)

# construct spline
splines = bspline(c(0, 4 * pi), nknots = 20)
# prepare/enhance data
obj = prepare(data, x, splines)
obj = enhance(obj)
sandmod = hero(obj)
plot(sandmod, ylim = range(data), lty = 2)
lines(x, data, col = "lightgrey")
lines(x, mu)
legend("bottomleft",
       legend = c("smoothed", "true", "observed"),
       lty = c(2, 1, 1),
       col = c("black", "black", "grey"))