QCSIS: Quantile Correlation-Sure Independence Screening (QC-SIS)

In QCSIS: Sure Independence Screening via Quantile Correlation and Composite Quantile Correlation

Description

The function implemrnts the quantile correlation-sure independence screening (QC-SIS).

Usage

QCSIS(x, y, tau, d)

Arguments

x	The design matrix, of dimensions n * p, without an intercept.
y	The response vector of dimension n * 1.
tau	The quantile(s) to be estimated. By default, tau=1:(n-1)/n.
d	The tuning parameter used to covarites had significant effect on the response variable, such as [n/log(n)],or n-1

Value

w	The estimate of w.
M	The subscript of x recuited by QC-SIS.

Author(s)

Xuejun Ma, Jingxiao Zhang, Jingke Zhou

References

Xuejun Ma and Jingxiao Zhang. Robust model-free feature screening via quantile correlation. Journal of Multivariate Analysis. Online, 2015.

Examples

n <- 20
p <- 200
r <- 0.05
x <- matrix(rnorm(n * p), n, p)
e <- rnorm(n, 0, 1)
inde <- sample(n, r * n)
x[inde, 1] <- 2 * sqrt(rchisq(r * n, df = p))
y <- 5 * x[, 1] + 5 * x[, 2] + 5 * x[, 3] + e
d <- 19
fit.QCSIS <- QCSIS(x = x, y = y, d = d)
fit.QCSIS$M

Example output

 [1]   3   4  63   6  83   1 124 180  81 103  22 131  84  95  14   7 192 101 144

QCSIS documentation built on May 1, 2019, 9:17 p.m.