Skip to contents

Add a Spline Smoother

Source: R/splineX.R
splineCI.Rd

Fit a smoothing spline and optionally add confidence bands.

Usage

splineX(x, ...)

# Default S3 method
splineX(x, ...)

# S3 method for class 'formula'
splineX(formula, data, subset, na.action = na.omit, weights, ...)

# S3 method for class 'SplineX'
lines(
  x,
  col = pal()[1],
  lwd = 2,
  lty = "solid",
  type = "l",
  band = list(conf.level = 0.95),
  ...
)

Arguments

x

spline object returned by splineX().

...

for splineX(), further arguments passed to stats::smooth.spline(); for lines(), further graphical parameters passed to lines().

formula

A formula of the form lhs ~ rhs, where lhs gives the response values and rhs the corresponding groups or explanatory variables.

data

an optional matrix or data frame (or similar; see stats::model.frame()) containing the variables in the formula. By default the variables are taken from environment(formula).

subset

an optional vector specifying a subset of observations to be used in the analysis.

na.action

A function which indicates what should happen when the data contain NAs. Defaults to getOption("na.action").

weights

optional vector of weights, a column of data or a vector of the length of the original data. It is subject to subset and na.action like the variables of the formula.

col

line color of the smoother.

lwd

line width.

lty

line type.

type

plotting type passed to lines().

band

controls the confidence band. May be TRUE, FALSE, NULL, NA, or a named list. The confidence level is specified via conf.level, all other elements are graphical parameters of the band. Default is list(conf.level = 0.95).

Value

splineX() returns the fit of stats::smooth.spline() with the additional class "SplineX". The lines() method returns no value (invisible NULL); it is called for its side effect of adding the smoother and its confidence band to the current plot.

Details

Confidence bands are controlled via band. These arguments can be:

  • FALSE, NULL or NA: suppress the band

  • TRUE: draw the band with default settings

  • a named list: customize the band appearance and confidence level

See also

loess(), scatter.smooth()

Other graphics.trendlines: lines.lm(), lines.loess()

Examples

op <- par(no.readonly = TRUE)
par(mfrow = c(1, 2))

x <- runif(100)
y <- rnorm(100)

plot(x, y)
lines(splineX(y ~ x))

plot(dist ~ speed, cars)
lines(splineX(dist ~ speed, cars))


# subset and weights, both evaluated in data
plot(dist ~ speed, cars)
lines(splineX(dist ~ speed, cars, subset = speed > 10,
              weights = sqrt(speed)))

plot(dist ~ speed, cars)
lines(
  splineX(dist ~ speed, cars),
  band = list(
    conf.level = 0.99,
    col = addOpacity("red", 0.3),
    border = "black"
  )
)


par(op)