Generic sparse group lasso cross validation using multiple possessors
sgl_cv(module_name, PACKAGE, data, parameterGrouping = NULL,
groupWeights = NULL, parameterWeights = NULL, alpha, lambda,
d = 100, compute_lambda = length(lambda) == 1, fold = 2,
sampleGroups = NULL, cv.indices = list(), responses = NULL,
max.threads = NULL, use_parallel = FALSE,
algorithm.config = sgl.standard.config)reference to objective specific C++ routines.
name of the calling package.
a list of data objects -- will be parsed to the specified module.
grouping of parameters, a vector of length \(p\). Each element of the vector specifying the group of the parameters in the corresponding column of \(\beta\).
the group weights, a vector of length length(unique(parameterGrouping)) (the number of groups).
a matrix of size \(q \times p\).
the \(\alpha\) value 0 for group lasso, 1 for lasso, between 0 and 1 gives a sparse group lasso penalty.
lambda.min relative to lambda.max (if compute_lambda = TRUE) or the lambda sequence for the regularization path,
a vector or a list of vectors (of the same length) with the lambda sequence for the subsamples.
length of lambda sequence (ignored if compute_lambda = FALSE)
should the lambda sequence be computed
the fold of the cross validation, an integer larger than \(1\) and less than \(N+1\).
Ignored if cv.indices != NULL.
If fold\(\le\)max(table(classes)) then the data will be split into fold disjoint subsets keeping the ration of classes approximately equal.
Otherwise the data will be split into fold disjoint subsets without keeping the ration fixed.
grouping of samples, the algorithm computing the cv.indices will try to equally divide the groups among the subsamples.
a list of indices of a cross validation splitting.
If cv.indices = NULL then a random splitting will be generated using the fold argument.
a vector of responses to simplify and return (if NULL (deafult) no formating will be done)
Deprecated (will be removed in 2018),
instead use use_parallel = TRUE and registre parallel backend (see package 'doParallel').
The maximal number of threads to be used.
If TRUE the foreach loop will use %dopar%.
The user must registre the parallel backend.
the algorithm configuration to be used.
the response, that is the y object in data as created by create.sgldata.
content will depend on the C++ response class
the cross validation splitting used
number of features used in the models
number of parameters used in the models
the lambda sequence used.