Sl.repara

INTERNAL routine to apply re-parameterization from log-determinant of penalty matrix, <code>ldetS</code> to
model matrix, <code>X</code>, blockwise.

Generalized additive (mixed) models, some of their extensions and
other generalized ridge regression with multiple smoothing
parameter estimation by (Restricted) Marginal Likelihood,
Generalized Cross Validation and similar, or using iterated
nested Laplace approximation for fully Bayesian inference. See
Wood (2017) <doi:10.1201/9781315370279> for an overview.
Includes a gam() function, a wide variety of smoothers, 'JAGS'
support and distributions beyond the exponential family.

Simon Wood

mgcv

Mixed GAM Computation Vehicle with Automatic Smoothness
Estimation

Sl.repara function

<dl><dt>rp</dt>
<dd>reparametrization.</dd>
<dt>X</dt>
<dd>if <code>X</code> is a matrix it is assumed to be a model matrix 
whereas if <code>X</code> is a vector it is assumed to be a parameter vector.</dd>
<dt>inverse</dt>
<dd>if <code>TRUE</code> an inverse re-parametrization is performed.</dd>
<dt>both.sides</dt>
<dd>if <code>inverse==TRUE</code> and <code>both.sides==FALSE</code> then
the re-parametrization only applied to rhs, as appropriate for a choleski factor.
If <code>both.sides==FALSE</code>, <code>X</code> is a vector and <code>inverse==FALSE</code> then <code>X</code> is
taken as a coefficient vector (so re-parametrization is inverse of that for the model matrix).</dd></dl>

Arguments

Simon N. Wood &lt;simon.wood@r-project.org&gt;.

Author

Applying re-parameterization from log-determinant of penalty matrix to
model matrix. — Sl.repara

<dl>

<dt>rp</dt>
<dd>reparametrization.</dd>


<dt>X</dt>
<dd>if <code>X</code> is a matrix it is assumed to be a model matrix 
whereas if <code>X</code> is a vector it is assumed to be a parameter vector.</dd>


<dt>inverse</dt>
<dd>if <code>TRUE</code> an inverse re-parametrization is performed.</dd>


<dt>both.sides</dt>
<dd>if <code>inverse==TRUE</code> and <code>both.sides==FALSE</code> then
the re-parametrization only applied to rhs, as appropriate for a choleski factor.
If <code>both.sides==FALSE</code>, <code>X</code> is a vector and <code>inverse==FALSE</code> then <code>X</code> is
taken as a coefficient vector (so re-parametrization is inverse of that for the model matrix).</dd>

</dl>

Sl.repara: Applying re-parameterization from log-determinant of penalty matrix to model matrix.

Description

Usage

Value

Arguments

Author