| Version: | 0.1-11 |
| Date: | 2019-05-09 |
| Title: | Robust Regression Functions |
| Author: | Ian M. Johnson <ijca2013@gmail.com> |
| Maintainer: | Ian M. Johnson <ijca2013@gmail.com> |
| Depends: | R (≥ 3.6.0) |
| Description: | Linear regression functions using Huber and bisquare psi functions. Optimal weights are calculated using IRLS algorithm. |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| Imports: | stats (≥ 3.6.0), Matrix (≥ 1.1.0), Rcpp (≥ 0.11.3) |
| LinkingTo: | Rcpp, RcppArmadillo |
| NeedsCompilation: | yes |
| Packaged: | 2019-05-09 23:07:35 UTC; ianjohns |
| Repository: | CRAN |
| Date/Publication: | 2019-05-10 07:50:03 UTC |
Predict y from X and b
Description
Predict y vector from X design matrix and b vector
Usage
fit_rcpp(X,b)
Arguments
X |
Design matrix |
b |
Estimates of beta |
Author(s)
Ian M. Johnson
Examples
j <- rep(1, 5)
x1 <- rnorm(5)
x2 <- rnorm(5, 10, 20)
X = as.matrix(data.frame(j, x1, x2))
b <- 1:3
fit_rcpp(X, b)
Median Absolute Deviation (MAD)
Description
Rcpp fast implementation of median absolute deviation (MAD)
Usage
mad_rcpp(r,scale_factor = 1.4826)
Arguments
r |
A numeric vector |
scale_factor |
Scale factor |
Author(s)
Ian M. Johnson
Examples
mad(1:100)
Median
Description
Rcpp fast implementation of median
Usage
median_rcpp(x)
Arguments
x |
A numeric vector containing the values whose median is to be computed. |
Author(s)
Ian M. Johnson
Examples
median_rcpp(1:100)
Tukey's Bisquare Psi Function
Description
Rcpp fast implementation of Tukey's Bisquare psi function
Usage
psiBS_rcpp(r,c)
Arguments
r |
A numeric vector |
c |
Tuning constant |
Author(s)
Ian M. Johnson
Examples
## Not run:
psiBS_rcpp(r,c)
## End(Not run)
Huber Psi Function
Description
Rcpp fast implementation of Huber's Psi Function
Usage
psiHuber_rcpp(r,c)
Arguments
r |
A numeric vector |
c |
Tuning constant |
Author(s)
Ian M. Johnson
Examples
## Not run:
psiHuber_rcpp(r,c)
## End(Not run)
Robust Fitting of Linear Models using Bisquare Psi Function
Description
Using iteratively reweighted least squares (IRLS), the function calculates the optimal weights to perform m-estimator or bounded influence regression. Returns robust beta estimates, mean squared error (MSE) and prints robust ANOVA table.
Usage
robustRegBS(formula,data,tune=4.685,m=TRUE,max.it=1000,tol=1e-5,anova.table=FALSE)Arguments
formula |
Model |
data |
A data frame containing the variables in the model. |
tune |
Tuning Constant. Default value of 4.685 is 95% asymptotically efficient against outliers |
m |
If |
max.it |
Maximum number of iterations to achieve convergence in IRLS algorithm |
tol |
Tolerance level in determining convergence |
anova.table |
If |
Details
M-estimates of beta should be used when evaluating least squares estimates of beta and diagnostics show outliers. Least squares estimates of beta should be used as starting points to achieve convergence.
Bounded influence estimates of beta should be used when evaluating least squares estimates of beta and diagnostics show large values of the "Hat Matrix" diagonals and outliers.
Note
Original package written in 2006
Author(s)
Ian M. Johnson
References
Tukey,
Birch, Robust F-Test, 1983
See Also
robustRegH()
Examples
data(stackloss)
robustRegBS(stack.loss~Air.Flow+Water.Temp,data=stackloss)
#If X matrix contained large values of H matrix (high influence points)
robustRegBS(stack.loss~Air.Flow+Water.Temp,data=stackloss,m=FALSE)
Robust Fitting of Linear Models using Huber Psi Function
Description
Using iteratively reweighted least squares (IRLS), the function calculates the optimal weights to perform m-estimator or bounded influence regression. Returns robust beta estimates, mean squared error (MSE) and prints robust ANOVA table
Usage
robustRegH(formula,data,tune=1.345,m=TRUE,max.it=1000,tol=1e-5,anova.table=FALSE)Arguments
formula |
Model |
data |
A data frame containing the variables in the model. |
tune |
Tuning Constant. Default value of 1.345 is 95% asymptotically efficient against outliers |
m |
If |
max.it |
Maximum number of iterations to achieve convergence in IRLS algorithm |
tol |
Tolerance level in determining convergence |
anova.table |
If |
Details
M-estimates of beta should be used when evaluating least squares estimates of beta and diagnostics show outliers. Least squares estimates of beta are used as starting points to achieve convergence.
Bounded influence estimates of beta should be used when evaluating least squares estimates of beta and diagnostics show large values of the "Hat Matrix" diagonals and outliers.
Note
Original package written in 2006
Author(s)
Ian M. Johnson
References
P. J. Huber (1981) Robust Statistics. Wiley.
Birch (1983) Robust F-Test
See Also
robustRegBS()
Examples
data(stackloss)
robustRegH(stack.loss~Air.Flow+Water.Temp,data=stackloss)
#If X matrix contained large values of H matrix (high influence points)
robustRegH(stack.loss~Air.Flow+Water.Temp,data=stackloss,m=FALSE)