Description Usage Arguments Details Functions See Also
View source: R/utility_functions.R
functions used to calculate cross validation error and used by
the cv.shim
function
1 2 3 4 5 6 7 | cv_lspath(outlist, x, y, foldid, nlambda, nlambda.beta, nlambda.gamma)
cvcompute(mat, foldid, nlams)
getmin_type(lambda, cvm, cvsd, type)
lambda.interp(lambda, s)
|
outlist |
list of cross validated fitted models. List is of length equal
to |
x |
Design matrix of dimension |
y |
response variable (matrix form) of dimension |
foldid |
numeric vector indicating which fold each observation belongs to |
nlambda |
total number of tuning parameter pairs. This includes those pairs of tuning parameters that didn't converge in the CV folds. |
nlambda.beta |
number of tuning paramters for the main effect |
nlambda.gamma |
number of tuning parameters for the interaction effects |
type |
should be one of "beta" or "gamma" |
outlist |
list containing results from cv.shim function. each element of the list is a run of shim_multiple_faster for each CV fold |
The output of the cv_lspath
function only returns values for those tuning
paramters that DID converge
cvcompute
: Computations for crossvalidation error
getmin_type
: Function that returns the tuning paramter corresponding
to the minimum cross validated error and cross validated error within 1
standard error of the minimum
lambda.interp
: Interpolation function. Currently not implemented.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.