plot.abesspca function

Creat plot from a fitted "abess" object

Produces a coefficient/deviance/tuning-value plot for a fitted "abess" object.

## S3 method for class 'abesspca'
plot(x, type = c("pev", "coef", "tune"), label = FALSE, ...)

Arguments

• x: A "abess" object.
• type: The type of terms to be plot in the y-axis. One of the following: "pev" (i.e., percent of explained variance), "coef" (i.e., coefficients), and "tune" (i.e., tuning value). Default is "coef".
• label: A logical value. If label = TRUE (the default), label the curves with variable sequence numbers.
• ...: Other graphical parameters to plot

Returns

No return value, called for side effects.

Note

If family = "mgaussian" or family = "multinomial", a coefficient plot is produced for each dimension of multivariate response.

Examples

abess_fit <- abesspca(USArrests, support.size = 1:4, sparse.type = "kpc")
plot(abess_fit)
plot(abess_fit, type = "coef")
plot(abess_fit, type = "tune")

See Also

print.abesspca, coef.abesspca, plot.abesspca.