buildMixtureDomains: Create RHugin domains for computation.

Description

The function is intended for internal use in DNAmixture, and it creates one network per marker.

IMPORTANT: This is the DNAmixturesLite package, which is intended as a service to enable users to try DNAmixtures without purchasing a commercial licence for Hugin. When at all possible, we strongly recommend the use of DNAmixtures rather than this lite-version. See https://dnamixtures.r-forge.r-project.org/ for details on both packages.

While the lite-version seeks to provide the full functionality of DNAmixtures, note that computations are much less efficient and that there are some differences in available functionality. Be aware that the present documentation is copied from DNAmixtures and thus may not accurately describe the implementation of this lite-version.

Usage

buildMixtureDomains(
  n.unknown,
  data,
  domainnamelist,
  write,
  dir = NULL,
  ntraces = 1,
  triangulate = TRUE,
  compile = TRUE,
  compress = TRUE,
  use.order = TRUE
)

Value

A list containing a list of pointers for RHugin domains. Additionally a list of the total size of the compressed junction tree in percent of the uncompressed size.

Arguments

n.unknown: Number of unknown contributors
data: A list of data.frames on the form returned by DNAmixtureData.
domainnamelist: List of names for networkfiles.
write: [not available in lite-version] Save networkfiles? Defaults to FALSE.
dir: Path to the networkfiles.
ntraces: Number of traces (EPG's)
triangulate: Triangulate the networks? Default is to triangulate the network using a good elimination order.
compile: Compile the networks? Defaults to TRUE.
compress: Compress the network? Defaults to TRUE and is strongly recommended for models with a large number of contributors.
use.order: Should the default elimination order be used for triangulation? Otherwise the "total.weight" heuristic for triangulation in HUGIN is used.

Author

Therese Graversen

Details

For one marker, the network contains variables

n_i_a: The number of alleles \(a\) that unknown contributor Ui possesses.
S_i_a: Cumulative allele counts, summarising how many alleles amongst types 1, ..., \(a\) that contributor \(i\) possesses.
O_r_a: Binary variable representing observed peak height for allele \(a\) in EPG \(r\). See setCPT.O for details.
D_r_a: Binary variable representing the event of a peak for allele \(a\) falling below threshold in EPG \(r\).
Q_r_a: Binary variable representing the event of a peak for allele \(a\) being smaller than the peak observed in EPG \(r\).

The network is by default triangulated, compiled, compressed, and, optionally, saved as a hkb-file. If the networks are large -- if there are many unknown contributors -- it is worth considering saving the networks rather than re-building them every time DNAmixture is called.