Compute actor size adjustment on a two way table

Description

When two objects are related through a bipartite graph it is sometimes appropriate to carry out special adjustments. One of the adjustments is called actor size adjustment. In this case the counts are adjusted according to how often the objects are referenced.

Usage

1
actorAdjTable(twT, eps = 1e-08)

Arguments

twT

A two way table as produced by twowayTable.

eps

A small quantity used to assess approximate equality.

Details

When testing for associations between entities, the social networks literature has developed a number of tools to help measure such associations. We can think of genes (actors) as being joined by citation in papers (events) and having two genes cited in the same paper (equivalent to two actors attending the same event) suggests that they are related to each other. However, some genes are cited in many papers and so we might want to discount the level of importance, as compared to genes that are cited less often. And additionally, some papers cite very many genes, and hence typically say less about them than a paper that cites rather fewer genes.

Value

An adjusted two way table, with elements named u11, u12, u21 and u22.

Author(s)

R. Gentleman

References

Testing Gene Associations Using Co-citation, by B. Ding and R. Gentleman. Bioconductor Technical Report, 2004

See Also

paperLen, twowayTable

Examples

1
2
  tw1 = twowayTable("10", "100", FALSE)
  actorAdjTable(tw1)