cvpocre {POCRE} | R Documentation |
Use k-Fold Cross-Validation to Choose the Tuning Parameter for POCRE
Description
Choose the optimal tuning parameter via k-fold cross-validation for POCRE.
Usage
cvpocre(y, x, n.folds=10, delta=0.1, maxvar=dim(x)[1]/2,
ptype=c('ebtz','ebt','l1','scad','mcp'), maxit=100,
maxcmp=10, gamma=3.7, lambda.init=1, tol=1e-6,
crit=c('press','Pearson','Spearman','Kendall'))
Arguments
y |
n*q matrix, values of q response variables (allow for multiple response variables). |
x |
n*p matrix, values of p predicting variables (excluding the intercept). |
n.folds |
number of folds to split the data (10-fold CV by default). |
delta |
step size of different values of the tuning parameter. |
maxvar |
maximum number of selected variables. |
ptype |
a character to indicate the type of penalty: |
maxit |
maximum number of iterations to be allowed. |
maxcmp |
maximum number of components to be constructed. |
gamma |
a parameter used by SCAD and MCP (=3.7 by default). |
lambda.init |
initial value of the tuning parameter (=1 by default). |
tol |
tolerance of precision in iterations. |
crit |
a character to indicate the validation criterion: |
Details
Use k-folds cross-validation to find the optinal value for the tuning parameter. The validation criterion can be chosen from PRESS, or different types of correlation coefficients, such as Pearson's, Spearman's, or Kendall's.
Value
The optimal value of the tuning parameter.
Author(s)
Dabao Zhang, Zhongli Jiang, Zeyu Zhang, Department of Statistics, Purdue University
References
Fan J and Li R (2001). Variable selection via nonconcave penalized likelihood and its oracle properties. Journal of the American Statistical Association, 96:1348-1360
Johnstone IM and Silverman BW (2004). Needles and straw in haystacks: empirical Bayes estimates of possibly sparse sequences. Annals of Statistics, 32: 1594-1649.
Zhang C-H (2010). Nearly unbiased variable selection under minimax concave penalty. The Annals of Statistics, 38: 894-942.
Zhang D, Lin Y, and Zhang M (2009). Penalized orthogonal-components regression for large p small n data. Electronic Journal of Statistics, 3: 781-796.
See Also
pocrescreen
, pocrepath
, pocre
.
Examples
## Not run:
data(simdata)
n <- dim(simdata)[1]
xx <- simdata[,-1]
yy <- simdata[,1]
# tp <- cvpocre(yy,xx,delta=0.01)
tp <- cvpocre(yy,xx)
print(paste(" pocre: Optimal Tuning Parameter = ", tp))
cvpres <- pocre(yy,xx,lambda=tp,maxvar=n/log(n))
## End(Not run)