RXshrink (version 1.4.3)

unr.biv: Specify pairs of GRR Coefficient Estimates for display in Bivariate Confidence Regions

Description

This function specifies which Pair of GRR estimates to display and the single (or dual) Confidence Level(s) of the Ellipse(s) displayed. Reguested confidence levels must both be equal to or greater than 0.05 and less than or equal to 0.95.

Usage

unr.biv(uraug, x1 = 1, x2 = 2, conf1 = 0.95, conf2 = 0.50)

Arguments

uraug

An output list object of class "unr.aug" for a GRR model with p >= 2 x-Variables.

x1

Integer index value >= 1 and <= p for the x-Coefficient displayed on the horizontal axis.

x2

Integer index value >= 1 and <= p for the x-Coefficient displayed on the vertical axis. Index x2 must differ from x1 to display a plot.

conf1

This first Confidence level must be >= 0.05 and <= 0.95 to display an Ellipse.

conf2

When the second Confidence level is >= 0.05 and <= 0.95, its Ellipse is displayed. No plot is displayed when both conf1 and conf2 are outside of the [0.05, 0.95] range.

Value

An output list object of class "unr.biv"...

p

Number of regression predictor variables.

LMobj

The lm() output object for the model fitted using unr.ridge().

bstar

The p by p+2 matrix of shrunken GRR coefficients. The first p correspond to "knots" on piecewise linear splines, the next to last contains minimum MSE risk estimates, and the last represents the shrinkge terminus.

mcal

p+2 increasing measures of shrinkage "Extent". The first value is 0 for the OLS = BLU Estimate, plotted as a Point of color "blue". The next to last corresponds to the "Extent" most likely to yield minimum MSE risk is colored "purple", and the last may be outside the plot frame.

ellip1

matrix[100, 2] of points on confidence ellipse 1.

conf1

confidence level of ellipse 1 within [0.05, 0.95].

ecor1

Pearson correlation between x1 and x2 in ellipse 1.

ellip2

matrix[100, 2] of points on confidence ellipse 2.

conf2

confidence level of ellipse 2 within [0.05, 0.95].

ecor2

Pearson correlation between x1 and x2 in ellipse 2.

References

Obenchain RL. (1977) Classical F-Tests and Confidence Regions for Ridge Regression. Technometrics 19, 429-439. http://doi.org/10.1080/00401706.1977.10489582

Murdoch DJ. and Chow ED. (1996). A graphical display of large correlation matrices. The American Statistician 50, 178-180.

Murdoch DJ. ellipse: Functions for Drawing Ellipses and Ellipse-Like Confidence Regions. https://CRAN.R-project.org/package=ellipse

See Also

unr.aug, ellipse.lm, unr.ridge