plot.amlps: Plot smooth functions of an additive model object.

In blapsr: Bayesian Inference with Laplace Approximations and P-Splines

Plot smooth functions of an additive model object.

Description

Displays a plot of the fitted additive smooth components of an amlps object. The routine can also be used to print a table of point and set estimates of an additive smooth term for a user-specified grid of values.

Usage

## S3 method for class 'amlps'
plot(x, xp, smoo.index, cred.int = 0.95, plot.cred = TRUE,
     np = 100, fit.col = "blue", shade.col = "gray75", show.plot = TRUE,
     show.info = TRUE, ...)

Arguments

x	An object of class amlps.
xp	A numeric vector of grid values on which to compute a point estimate and pointwise credible interval for the smooth function specified in smoo.index. The components of xp must be within the range of the observed covariate values for the corresponding smooth function. Results will be displayed in a table.
smoo.index	The index of the smooth function. For instance smoo.index = 2 refers to the second smooth function specified in the formula of the amlps routine.
cred.int	The level of the pointwise credible interval to be computed for the smooth additive term. Default is 0.95.
plot.cred	Logical. Should the credible intervals be plotted? Default is TRUE.
np	The number of points used to construct the plot of the smooth additive function. Default is 100 and allowed values are between 20 and 200.
fit.col	The color of the fitted curve.
shade.col	The shading color for the credible intervals.
show.plot	Logical. Should the plot be displayed? Default is TRUE.
show.info	Logical. Should the table of point and set estimates of the smooth function on the specified xp values be displayed? Default is TRUE.
...	Further arguments to be passed to plot.

Details

Produces a plot of a smooth additive term fitted with the amlps function. On the y-axis, the estimated effective dimension of the smooth term is also displayed. At the bottom of each plot, vertical ticks indicate the location of the covariate values. The labels on the x-axis correspond to the covariate name associated to the smooth term.

Value

If xp is unspecified (the default), the routine will only return a plot of the estimated smooth curve. Otherwise, it provides a list with the following components:

xp	The chosen points on which to compute the smooth fit.
sm.xp	The estimated smooth fit at points specified in xp.
sm.low	The lower bound of the pointwise credible interval for the smooth additive function at points specified in xp.
sm.up	The upper bound of the pointwise credible interval for the smooth additive function at points specified in xp.
cred.int	The chosen level to compute credible intervals.
smoo.index	The index of the smooth function.

Author(s)

Oswaldo Gressani [e-mail].

See Also

amlps, amlps.object, print.amlps

Examples

### Classic simulated data example

set.seed(3)

sim.data <- simgamdata(setting = 2, n = 200, dist = "gaussian", scale = 0.3)
plot(sim.data)         # Scatter plot of response
data <- sim.data$data  # Simulated data frame

# Fit model
fit <- amlps(y ~ z1 + z2 + sm(x1) + sm(x2), data = data, K = 20)
fit

# Plot fit of second function and results for a specific grid x
plot(fit, xp = c(-0.8, -0.4, 0, 0.4, 0.8), smoo.index = 2, ylim=c(-3, 3))

blapsr documentation built on Aug. 20, 2022, 5:05 p.m.