Qtools (version 1.0)

rrq: Restricted Regression Quantiles

Description

This function fits a restricted quantile regression model to avoid crossing of quantile curves.

Usage

rrq(formula, tau, data, subset, weights, na.action, method = "fn",
model = TRUE, contrasts = NULL, ...)
rrq.fit(x, y, tau, method = "fn", ...)
rrq.wfit(x, y, tau, weights, method = "fn", ...)

References

He X. Quantile curves without crossing. The American Statistician 1997;51(2):186-192. Koenker R (2013). quantreg: Quantile Regression. R package version 5.05. URL http://CRAN.R-project.org/package=quantreg.

Examples

data(esterase)

# Fit standard quantile regression
fit <- quantreg::rq(Count ~ Esterase, data = esterase, tau = c(.1,.25,.5,.75,.9))
yhat <- fit$fitted.values

# Fit restricted quantile regression
fitr <- rrq(Count ~ Esterase, data = esterase, tau = c(.1,.25,.5,.75,.9))
yhat2 <- predict(fitr)


# Plot results
par(mfrow = c(1, 2))

# Plot regression quantiles
with(esterase, plot(Count ~ Esterase, pch = 16, cex = .8))
apply(yhat, 2, function(y,x) lines(x,y,lwd = 1.5), x = esterase$Esterase)

# Plot restricted regression quantiles
with(esterase, plot(Count ~ Esterase, pch = 16, cex = .8))
apply(yhat2, 2, function(y,x) lines(x,y,lwd = 1.5), x = esterase$Esterase)