naginterfaces.library.correg.lmm_​init_​combine

naginterfaces.library.correg.lmm_init_combine(hlmm, xcomm, ycomm)

lmm_init_combine combines output from multiple calls to lmm_init() or mixeff_hier_init().

For full information please refer to the NAG Library document for g02jg

https://support.nag.com/numeric/nl/nagdoc_30/flhtml/g02/g02jgf.html

Parameters

hlmmHandle

If the two sets of communication arrays were generated using lmm_init(), then a G22 handle as generated by one of the calls to lmm_init().

If the two sets of communication arrays were generated using mixeff_hier_init(), then this argument is not referenced and need not be set.

xcommdict, communication object, modified in place

Communication structure for $D_x$.

This argument must have been initialized by a prior call to lmm_init() or mixeff_hier_init().

ycommdict, communication object

Communication structure for $D_y$.

This argument must have been initialized by a prior call to lmm_init() or mixeff_hier_init().

Raises

NagValueError

(errno $11$)

$hlmm$ has not been initialized or is corrupt.

(errno $12$)

$hlmm$ is not a G22 handle as generated by lmm_init().

(errno $41$)

On entry, $xcomm$[‘iopts’] has not been initialized correctly.

(errno $42$)

On entry, the information stored in $xcomm$[‘iopts’] and $ycomm$[‘iopts’] is not compatible.

(errno $71$)

On entry, $ycomm$[‘iopts’] has not been initialized correctly.

Notes

Let $D_x$ and $D_y$ denote two sets of data, each with $m$ variables and $n_x$ and $n_y$ observations respectively. Let $C_x$ and $C_y$ denote two sets of communication arrays constructed by lmm_init() or mixeff_hier_init() from datasets $D_x$ and $D_y$ respectively. Then, given $C_x$ and $C_y$, lmm_init_combine constructs a set of communication arrays, $C_w$, as if a dataset $D_z$, with $m$ variables and $n_x+n_y$ observations were supplied to lmm_init() or mixeff_hier_init(), with $D_z$ constructed as

$$\begin{array}{c}{D}_z=\left(\begin{array}{c}{D}_x\\ D_y\end{array}\right)\text{.}\end{array}$$

Splitting, and then recombining, the data in this manner allows for datasets with an arbitrarily large number of observations ($n$) to be analysed and the preprocessing routines, lmm_init() or mixeff_hier_init(), to be run in parallel.

It should be noted that, while the information in $C_z$, should be consistent with the information in the communication arrays obtained by supplying $D_z$ to lmm_init() or mixeff_hier_init(), the ordering of that information may change. In practice, this means that whilst an analysis run using a set of communication arrays constructed using lmm_init_combine should give similar results to an analysis run using a set of communication arrays constructed directly from lmm_init() or mixeff_hier_init() they will not necessarily be identical. In addition, the order of the parameter estimates, $\nu$ and $\beta$ may differ.

See also

naginterfaces.library.examples.correg.lmm_init_combine_ex.main()