mgcv_lambda

penalties

List of <code>nxn</code> matrices. Usually output of <code><a rd-options="" href="/link/createXXblocks?package=multiridge&version=1.11" data-mini-rdoc="multiridge::createXXblocks">createXXblocks</a></code>.

XXblocks

Response vector: numeric, binary, factor or <code>survival</code>.

Character. Any of <code>c("linear", "logistic", "cox")</code>. Is inferred from
<code>Y</code> when <code>NULL</code>.

model

Boolean. Should the score be printed?

printscore

Numerical vector of length 3 or <code>NULL</code> when pairs are absent. Represents the indices (in <code>XXblocks</code>) of the two data blocks involved in pairing, plus the index of the paired block.

pairing

sigmasq

Boolean. Should the error variance be optimized as well? Only relevant for <code>model="linear"</code>.

opt.sigma

Computed maximum marginal likelihood score for given penalty parameters using <code>mgcv</code>.

Multi-penalty linear, logistic and cox ridge regression, including estimation of the penalty parameters by efficient (repeated) cross-validation and marginal likelihood maximization. Multiple high-dimensional data types that require penalization are allowed, as well as unpenalized variables. Paired and preferential data types can be specified. See Van de Wiel et al. (2021), <arXiv:2005.09301>.

Mark van de Wiel

multiridge

Fast Cross-Validation for Multi-Penalty Ridge Regression

mgcv_lambda function

List of <code>nxn</code> matrices. Usually output of <code><a rd-options='' href='createXXblocks'>createXXblocks</a></code>.

mgcv_lambda: Maximum marginal likelihood score

Description

Usage

Arguments

Value

Details

References

See Also