Collection of miscellaneous useful and semi-useful functions

Description

Collection of miscellaneous useful and semi-useful functions and add-on functions that enhances a number of existing packages and provides In particular in relation to statistical genetics

Details

how to use the package, including the most important ~~

Author(s)

Claus Thorn Ekstrøm ekstrom@sund.ku.dk
Maintainer: Claus Thorn Ekstrøm ekstrom@sund.ku.dk

References

Ekstrøm, C. (2017). The R Primer. 2nd edition, Chapman & Hall.

Create adjacency matrix

Description

Create an adjacency matrix from a set of nodes and edges.

Usage

adjacency(from, to, weight = 1, directed = TRUE)

Arguments

Value

Returns a sparse adjacency matrix

Author(s)

Claus Ekstrom ekstrom@sund.ku.dk

Examples

from <- c("A", "A", "A", "B", "C")
to <- c("B", "C", "D", "D", "E")
adjacency(from, to)

from <- c("A", "A", "A", "B", "C")
to <- c("B", "C", "D", "D", "E")
weights <- c(1, .5, 1, .7, 1)
adjacency(from, to, weights)

Hyperlink-induced topic search

Description

Hyperlink-induced topic search (HITS) is a link analysis algorithm that is also known as hubs and authorities. It rates nodes by comparing arrows pointing in and out of nodes in an asymmetrical graph.

Usage

hits(adjmatrix, maxiter = 100L, tol = 1e-05)

Arguments

Details

Hubs are nodes with a lot of arrows pointing out while authorities are node with a lot of arrows pointing in.

Value

Returns a list with three elements: authorities (a vector) of and hubs (a vector), and number of iterations used.

Author(s)

Claus Ekstrom ekstrom@sund.ku.dk

References

Kleinberg, Jon (1999). "Authoritative sources in a hyperlinked environment" (PDF). Journal of the ACM. 46 (5): 604–632. doi:10.1145/324133.324140

Examples

from <- c("A", "A", "A", "B", "C")
to <- c("B", "C", "D", "D", "E")
hits(adjacency(from, to))

Construct family id vector from pedigree trio information

Description

Create a vector of length n, giving the family id of each subject. If the pedigree is totally connected, then everyone will end up in tree 1, otherwise the tree numbers represent the disconnected subfamilies. Singleton subjects each have unique family numbers.

No check is done to ensure that the id, fid, and mid actually refer to proper family structure. References to ids in the fid and mid arguments that are not part of the id vector are considered founders.

Usage

make_family_id(id, fid, mid)

Arguments

Value

Returns an integer vector giving the family index

Author(s)

Claus Ekstrom ekstrom@sund.ku.dk

Examples

id <- 1:11
fid <- c(NA, NA, 1, 1, NA, 23, 45, 5, 5, 7, NA)
mid <- c(NA, NA, 2, 2, 65, NA, 46, 6, 6, 6, 0)
make_family_id(id, fid, mid)

Generate variables (or lists) of siblings from a file of ids of persons and their father and mother.

Description

The function generates for each person lists of maternal half-sibs, paternal half-sibs and full sibs. Optionally these are expanded to separate columns in a data.table.

Usage

mksib(obj, ns = 3, expand.vars = TRUE)

Arguments

Details

There are no checks of persons being both mother and father, nor being its own parent and incest checks are not performed. In other words, the obj is assumed to be sane, but possibly immoral.

Value

A data.table with the columns of the obj and columns for ns maternal, paternal and full sibs, named ms1, ms2, ... ps1, ps2, ... fs1, fs2.

If expand.vars=FALSE there will instead be three columns of lists named msibs, psibs and fsibs.

Author(s)

Claus Thorn Ekstrøm, ekstrom@sund.ku.dk, Bendix Carstensen, b@bxc.dk

Examples

library( data.table )
id <- 1:12
pid <- c(NA,  1, 1, 1, NA, 23, 45, 5, 5, 7, 12, NA)
mid <- c(NA, NA, 2, 2, 12, NA, 46, 6, 6, 6, NA, 12)
indd <- data.table( id, mid, pid )
indata <- copy( indd )
indata

str( xx <- mksib( indata ) )
xx

zz <- mksib( indata, 2, e=FALSE )
zz

Validate pedigree trio information consistency

Description

Simple tests to chech the consistency of the pedigree trio family data. Currently the following checks are undertaken: 1) that no duplicates ids are found; 2) that the primary id is not missing for anyone; 3) that founders have both the father and mother id missing; 4) that individuals are not both classified as male (fathers and mothers);

Usage

validate_trio_consistency(id, fid, mid, sex = NULL)

Arguments

Details

There are no checks of persons being both mother and father, nor being its own parent and incest checks are not performed. In other words, the obj is assumed to be sane, but possibly immoral.

Value

Throws an error if an inconsistency is found. Otherwise returns TRUE.

Author(s)

Claus Thorn Ekstrøm, ekstrom@sund.ku.dk

Examples

library("data.table")
id <- 1:12
fid <- c(NA,  0, 1, 1, NA, 23, 45, 5, 5, 7, 10, 10)
mid <- c(NA, NA, 2, 2,  0, 56, 46, 6, 6, 6, 9, 11)

validate_trio_consistency(id, fid, mid)