qrlinear: Quantile regression using R.

In YulieKing/StatComp18026: Example functions for 'Statistical Computing' course

Description

Quantile regression by Linear Programming using R. Which means turning Quantile regression into a linear programming problem

Usage

qrlinear(x, y, tau, Intercept = T)

Arguments

x	the observed value of independent variables x should be a vector or matrix.(numeric)
y	the observed value of dependent variable and y should be a vector. (numeric)
tau	the quantile level and tau should be in (0,1). (numeric)
Intercept	if TRUE then the reslut including the intercept coefficient (logical)

Value

the regression coefficients

Examples

## Not run: 
data(engel)
b1hat<-matrix(NA,ncol=2,nrow=6);b2hat<-matrix(NA,ncol=2,nrow=6)
taus <- c(.05,.1,.25,.75,.9,.95)
for(i in 1:length(taus)){
 b1hat[i,]=qrlinear(income,foodexp,taus[i],Intercept=T)
 b2hat[i,]=rq(foodexp~income,taus[i])$coef
 }
 print(b1hat);print(b2hat)

## End(Not run)

YulieKing/StatComp18026 documentation built on May 29, 2019, 8:34 a.m.