plot.bsrr {bestridge}

Produces a coefficient profile plot of the coefficient or loss function paths

Description

Produces a coefficient profile plot of the coefficient or loss function paths

Usage

## S3 method for class 'bsrr'
plot(
  x,
  type = c("tune", "coefficients"),
  lambda = NULL,
  sign.lambda = 0,
  breaks = T,
  K = NULL,
  ...
)

Arguments

x

A "bsrr" object.

type

One of "tune", "coefficients", "both". For "bsrr" with L_2 shrinkage: If (type = "tune"), the path of corresponding information criterion or cross-validation loss is provided; If type = "coefficients", a lambda should be provided and and this funciton provides a coefficient profile plot of the coefficient; For "bsrr" object without L_2 shrinkage: If type = "tune", a path of lcorresponding information criterion or cross-validation loss is provided. If type = "coefficients", it provides a coefficient profile plot of the coefficient.

lambda

For "bsrr" with L_2 shrinkage: To plot the change of coefficients with lambda equals this value for type = "coefficients" or type = "both".

sign.lambda

For "bsrr" with L_2 shrinkage: A logical value indicating whether to show lambda on log scale. Default is 0.

breaks

For "bsrr" object without L_2 shrinkage: If TRUE, a vertical line is drawn at a specified break point in the coefficient paths.

K

For "bsrr" object without L_2 shrinkage: Which break point should the vertical line be drawn at. Default is the optimal model size.

...

Other graphical parameters to plot

Value

No return value, called for plots generation

Author(s)

Liyuan Hu, Kangkang Jiang, Yanhang Zhang, Jin Zhu, Canhong Wen and Xueqin Wang.

See Also

bsrr.

Examples

# Generate simulated data
n <- 200
p <- 20
k <- 5
rho <- 0.4
seed <- 10
Tbeta <- rep(0, p)
Tbeta[1:k*floor(p/k):floor(p/k)] <- rep(1, k)
Data <- gen.data(n, p, k, rho, family = "gaussian", beta = Tbeta, seed = seed)
lambda.list <- exp(seq(log(5), log(0.1), length.out = 10))
lm.bsrr <- bsrr(Data$x, Data$y, method = "pgsection")

# generate plots
plot(lm.bsrr)