Learn R Programming

simest (version 0.4-1-1)

cvx.lse.reg: Convex Least Squares Regression

Description

This function provides an estimate of the non-parametric regression function with a shape constraint of convexity and no smoothness constraint. Note that convexity by itself provides some implicit smoothness.

Usage

cvx.lse.reg(t, z, w = NULL,...)
# S3 method for cvx.lse.reg
plot(x, diagnostics = TRUE,
     ylab = quote(y ~ "and" ~ hat(y) ~ " values"),
     pch = "*", cex = 1, lwd = 2, col2 = "red", ablty = 4, ...)
# S3 method for cvx.lse.reg
print(x, digits = getOption("digits"), ...)
# S3 method for cvx.lse.reg
predict(object, newdata = NULL, deriv = 0, ...)

Value

An object of class cvx.lse.reg, basically a list including the elements

x.values

sorted t values provided as input.

y.values

corresponding z values in input.

fit.values

corresponding fit values of same length as that of x.values.

deriv

corresponding values of the derivative of same length as that of x.values.

iter

number of steps taken to complete the iterations.

residuals

residuals obtained from the fit.

minvalue

minimum value of the objective function attained.

convergence

a numeric indicating the convergence of the code.

Arguments

t

a numeric vector giving the values of the predictor variable.

z

a numeric vector giving the values of the response variable.

w

an optional numeric vector of the same length as t; Defaults to all elements \(1/n\).

...

additional arguments.

diagnostics

for the plot() method; if true, as by default, produce diagnostics, notably residual plots additionally.

ylab, pch, cex, lwd, col2, ablty

further optional argument to the plot() method; the last two for the color and line type of some plot components.

digits

the number of significant digits, for numbers in the print() method.

x, object

an object of class "cvx.lse.reg".

newdata

a matrix of new data points in the predict function.

deriv

a numeric either 0 or 1 representing which derivative to evaluate.

Author

Arun Kumar Kuchibhotla

Details

The function minimizes $$\sum_{i=1}^n w_i(z_i - \theta_i)^2,$$ subject to $$\frac{\theta_2 - \theta_1}{t_2 - t_1}\le\cdots\le\frac{\theta_n - \theta_{n-1}}{t_n - t_{n-1}},$$ for sorted \(t\) values (and \(z\) permuted accordingly such that \((t_i, z_i)\) stay pairs.

This function previously used the coneA() function from the coneproj package to perform the constrained minimization of least squares. Currently, the code makes use of the nnls() function from package nnls for the same purpose.

The plot method provides a scatterplot along with the fitted curve; it also includes some diagnostic plots for residuals. The predict() method allows computation of the first derivative.

References

Chen, D. and Plemmons, R. J. (2009) Non-negativity Constraints in Numerical Analysis. Symposium on the Birth of Numerical Analysis.

Liao, X. and Meyer, M. C. (2014) coneproj: An R package for the primal or dual cone projections with routines for constrained regression. Journal of Statistical Software 61(12), 1--22.

Examples

Run this code
args(cvx.lse.reg)
x <- runif(50,-1,1)
y <- x^2 + rnorm(50,0,0.3)
cvxL <- cvx.lse.reg(x, y)
print(cvxL)
plot(cvxL)
predict(cvxL, newdata = rnorm(10,0,0.1))

Run the code above in your browser using DataLab