Qij_class: Constructor function for the class Qij

In eriqande/CKMRsim: Inference of Pairwise Relationships Using Likelihood Ratios

Constructor function for the class Qij

Description

The Qij class holds the output of simulated genotype pairs whose probabilities have been evaluated under different pairwise relationship hypotheses. It is output by the function simulate_Qij.

Usage

Qij_class(
  Q,
  unlinked,
  forceLinkagePO,
  miss_mask_mat,
  rando_miss_wts,
  rando_miss_n
)

Arguments

Q	the list of lists to be turned into a Qij object.
unlinked	Logical that says whether simulation was of unlinked markers or not
forceLinkagePO	Logical. If linked markers, should PO be forced to be simulated via Mendel
miss_mask_mat	A logical matrix with length(YL) columns and reps rows. The (r,c)-th is TRUE if the c-th locus should be considered missing in the r-th simulated sample. This type of specification lets the user simulate either a specific pattern of missingness, if desired, or to simulate patterns of missing data given missing data rates, etc.
rando_miss_wts	weights to be given to different loci that influence whether they will be one of the rando_miss_n missing loci in any iteration. These will be recycled (or truncated) to have length equal to the number of loci, and they will be normalized to sum to one as appropriate (so you can provide them in unnormalized form.) The idea of this is to be able to use observed rates of missingness amongst loci to mask some loci as missing. Given as a comma-delimited string in column "rando_miss_wts" in the output.
rando_miss_n	a single number less than the number of loci. Each iteration, rando_miss_n loci will be considered missing, according to the rando_miss_wts. This let's you get a sense for how well you will do, on average, with a certain number of missing loci.

Details

A Qij object is quite simply a list of lists of vectors. At the first level the components are named by relationships from which data were simulated (the "froms"). Within each of those is a named list of relationships for which the probability of each simulated genotype pair was evaluated (the "tos").

eriqande/CKMRsim documentation built on Aug. 2, 2024, 7:23 a.m.