Algorithmic constants and parameters for
 running <code>rrvglm</code> are set using this
 function.
 Doubly constrained RR-VGLMs (DRR-VGLMs) are
 also catered for.

rrvglm.control

An implementation of about 6 major classes of
statistical regression models. The central algorithm is
Fisher scoring and iterative reweighted least squares.
At the heart of this package are the vector generalized linear
and additive model (VGLM/VGAM) classes. VGLMs can be loosely
thought of as multivariate GLMs. VGAMs are data-driven
VGLMs that use smoothing. The book "Vector Generalized
Linear and Additive Models: With an Implementation in R"
(Yee, 2015) <DOI:10.1007/978-1-4939-2818-7> gives details of
the statistical framework and the package. Currently only
fixed-effects models are implemented. Many (100+) models and
distributions are estimated by maximum likelihood estimation
(MLE) or penalized MLE. The other classes are RR-VGLMs
(reduced-rank VGLMs), quadratic RR-VGLMs, doubly constrained
RR-VGLMs, quadratic RR-VGLMs, reduced-rank VGAMs,
RCIMs (row-column interaction models)---these classes perform
constrained and unconstrained quadratic ordination (CQO/UQO)
models in ecology, as well as constrained additive ordination
(CAO). Hauck-Donner effect detection is implemented.
Note that these functions are subject to change;
see the NEWS and ChangeLog files for latest changes.

Thomas Yee

VGAM

Vector Generalized Linear and Additive Models

Cleve Moler

rrvglm.control function

Control Function for rrvglm() — rrvglm.control

<dl>

 <dt>Rank</dt>
<dd>The numerical rank \(R\) of the
 model. Must be an element from the set
 {1,2,...,min(\(M\),p2)}.
 Here, the vector of explanatory
 variables x is partitioned into
 (x1, x2), which is of dimension
 p1+p2. The variables making
 up x1 are given by the terms in
 <code>noRRR</code> argument, and the rest of the
 terms comprise x2.

</dd>

 <dt>Corner</dt>
<dd>Logical indicating whether corner
 constraints are to be used.
 Strongly recommended as the only
 method for fitting RR-VGLMs and
 DRR-VGLMs.
 This is one
 method for ensuring a unique solution
 and the availability of standard errors.
 If <code>TRUE</code> then <code>Index.corner</code>
 specifies the position of the corner
 constraints.
 Also used by DRR-VGLMs, and this is
 known as RCCs (see below).

</dd>

 <dt>Index.corner</dt>
<dd>Specifies the \(R\) rows of
 the corner constraints, e.g., they hold an
 order-\(R\) identity matrix for RR-VGLMs.
 The default are the first \(R\) positive
 integers that exclude <code>str0</code>.

 

For DRR-VGLMs one needs
to have (restricted) corner constraints.
Then argument <code>str0</code> will be ignored.
If there are structural zeros then they
should be built into the constraint matrices
(see <code>H.A.thy</code>).

</dd>

 <dt>noRRR</dt>
<dd>Formula giving terms that are not
 to be included in the reduced-rank
 regression. That is, <code>noRRR</code>
 specifes which explanatory variables
 are in the x1 vector of
 <code>rrvglm</code>, and the rest go into
 x2.
 I will write x1 as \(x_1\)
 and x2 as \(x_2\) hereafter.
 The \(x_1\) variables
 constitute the \(\bold{B}_1\)
 matrix in Yee and Hastie (2003). Those
 \(x_2\) variables which are subject
 to the reduced-rank regression correspond
 to the \(\bold{B}_2\)
 matrix. Set <code>noRRR = NULL</code> for the
 reduced-rank regression to be applied to
 every explanatory variable including the
 intercept.

</dd>

 <dt>str0</dt>
<dd>Integer vector specifying which rows of the
 estimated constraint matrices (A)
 are to be all zeros. These are called
 structural zeros. Must not have
 any common value with <code>Index.corner</code>,
 and be a subset of the vector <code>1:M</code>.
 The default, <code>str0 = NULL</code>, means no
 structural zero rows at all.
 This argument is ignored by DRR-VGLMs.


</dd>

 <dt>Crow1positive</dt>
<dd>Currently this argument has no effect.
 In the future, it may be a
 logical vector of length <code>Rank</code>
 (recycled if necessary): are
 the elements of the first
 row of \(C\) positive?
 Because of corner constraints (<code>diag(Rank)</code>)
 this argument currently must be <code>NULL</code>.
 Having a logical vector would mean some of the diagonals
 might be \(1\) and others \(-1\).
























</dd>



 <dt>trace</dt>
<dd>Logical indicating if output should be
 produced for each iteration.




</dd>





 <dt>Bestof</dt>
<dd>Integer. The best of <code>Bestof</code>
 models fitted is returned. This argument
 helps guard against local solutions by
 (hopefully) finding the global solution
 from many fits. The argument works only
 when the function generates its own initial
 value for C, i.e., when C
 is not passed in as initial values.

</dd>

 <dt>H.A.thy, H.C</dt>
<dd>Lists.
 DRR-VGLMs are Doubly constrained
 RR-VGLMs where A has
 <code>Rank</code>
 constraint matrices
 (one for each column)
 in a list called <code>H.A.thy</code>,
 and C has a
 constraint matrix for each row,
 i.e., for each explanatory
 variable making up it.
 The class
 <code>"drrvglm"</code> may arise for such models.
 So <code>H.C</code>
 should be a named list of \(p_2\)
 constraint matrices,
 each one for a different row of C,
 i.e., \(p_2\) is the number of variables
 making up the latent variable.
 Note that if <code>H.C</code> has
 <code><a href='https://rdrr.io/r/base/names.html'>names</a></code> then
 matching is done with that,
 and the components of <code>H.C</code>
 are reordered if they are not sorted
 according to the terms in <code>formula</code>.
 If they are not named, then
 their order is used,
 for example, <code>H.C[[1]]</code>
 and <code>H.C[[2]]</code> are taken as
 the constraint matrices for the first two
 variables of the latent variable(s).







</dd>

 <dt>Algorithm</dt>
<dd>Character string indicating what algorithm is
 to be used. The default is the first one.
 The choice <code>"derivative"</code> has been
 withdrawn in VGAM 1.1-10. 

</dd>

 <dt>Ainit, Cinit</dt>
<dd>Initial A and C matrices which
 may speed up convergence. They must be of
 the correct dimension.

</dd>

 <dt>sd.Cinit</dt>
<dd>Standard deviation of the initial values
 for the elements of C.
 These are normally distributed with
 mean zero. This argument is used only if
 <code>Use.Init.Poisson.QO = FALSE</code>.

</dd>

 <dt>Etamat.colmax</dt>
<dd>Positive integer, no smaller than
 <code>Rank</code>. Controls the amount of
 memory used by <code>.Init.Poisson.QO()</code>.
 It is the maximum number of columns allowed
 for the pseudo-response and its weights.
 In general, the larger the value, the
 better the initial value. Used only if
 <code>Use.Init.Poisson.QO = TRUE</code>.

</dd>

</dl>




























































































































































<dl>

 <dt>Use.Init.Poisson.QO</dt>
<dd>Logical indicating whether the
 <code>.Init.Poisson.QO()</code> should be used
 to obtain initial values for the C.
 The function uses a new method that can
 work well if the data are Poisson counts
 coming from an equal-tolerances QRR-VGLM
 (CQO). This option is less realistic for
 RR-VGLMs compared to QRR-VGLMs.
</dd>

 <dt>checkwz</dt>
<dd>logical indicating whether the diagonal
 elements of the working weight matrices
 should be checked whether they are
 sufficiently positive, i.e., greater than
 <code>wzepsilon</code>. If not, any values less
 than <code>wzepsilon</code> are replaced with
 this value.

</dd>

 <dt>noWarning, Check.rank, Check.cm.rank</dt>
<dd>Same as <code>vglm.control</code>.
 Ignored for VGAM 0.9-7 and higher.


</dd>

 <dt>wzepsilon</dt>
<dd>Small positive number used to test whether
 the diagonals of the working weight matrices
 are sufficiently positive.
</dd>

</dl>























 
























































<dl>




 <dt>...</dt>
<dd>Variables in ... are passed into
 <code>vglm.control</code>.
 If the derivative algorithm is used
 then ... are also passed into
 <code>rrvglm.optim.control</code>;
 and if the alternating algorithm is
 used then ... are also passed into
 <code>valt0.control</code>.

</dd>

</dl>
In the above, \(R\) is the <code>Rank</code> and

\(M\) is the number of linear predictors.

rrvglm.control: Control Function for rrvglm()

Description

Usage

Arguments

Value

Details

References

See Also

Examples