msaenet (version 2.1)

msasnet: Multi-Step Adaptive SCAD-Net

Description

Multi-Step Adaptive SCAD-Net

Usage

msasnet(x, y, family = c("gaussian", "binomial", "poisson", "cox"), init = c("snet", "ridge"), nsteps = 2L, nfolds = 5L, gammas = c(2.01, 2.3, 3.7, 200), alphas = seq(0.05, 0.95, 0.05), eps = 1e-04, max.iter = 10000L, gamma = 1, seed = 1001, parallel = FALSE, verbose = FALSE)

Arguments

x

Data matrix.

y

Response vector if family is "gaussian", "binomial" or "poisson". If family is "cox", a response matrix made by Surv.

family

Model family, can be "gaussian", "binomial", "poisson", or "cox".

init

Type of the penalty used in the initial estimation step. Can be "snet" or "ridge".

nsteps

How many adaptive estimation steps? At least 2. We assume adaptive SCAD-net has only 1 adaptive estimation step.

nfolds

Fold numbers of cross-validation.

gammas

Vector of candidate gammas to use in SCAD-Net.

alphas

Vector of candidate alphas to use in SCAD-Net.

eps

Convergence threshhold to use in SCAD-net.

max.iter

Maximum number of iterations to use in SCAD-net.

gamma

Scaling factor for adaptive weights: weights = coefs^(-gamma).

seed

Two random seeds for cross-validation fold division in two estimation steps.

parallel

Logical. Enable parallel parameter tuning or not, default is FALSE. To enable parallel tuning, load the doParallel package and run registerDoParallel() with the number of CPU cores before calling this function.

verbose

Should we print out the estimation progress?

Value

List of coefficients beta and ncvreg model object model.

Examples

dat = msaenet.sim.gaussian(n = 150, p = 500, rho = 0.6,
                           coef = rep(1, 5), snr = 2, p.train = 0.7,
                           seed = 1001)

msasnet.fit = msasnet(dat$x.tr, dat$y.tr,
                      gammas = 3.7, alphas = seq(0.3, 0.9, 0.3),
                      nsteps = 3L, seed = 1003)

print(msasnet.fit)
msaenet.nzv(msasnet.fit)
msaenet.fp(msasnet.fit, 1:5)
msaenet.tp(msasnet.fit, 1:5)
msasnet.pred = predict(msasnet.fit, dat$x.te)
msaenet.rmse(dat$y.te, msasnet.pred)
plot(msasnet.fit)