Skip to contents

Compute skill co-occurrence for a single profession

Source: R/cooccurrence.R
dot-cooc_single.Rd

Builds a sparse term-document matrix for a given profession, computes the co-occurrence via Matrix::crossprod(), and extracts the top edges using an igraph largest-component filter. This is the workhorse behind cooc_all_professions().

Usage

.cooc_single(
  professione,
  data,
  group_col = "preferredLabel",
  top_n = 30L,
  min_weight = 10L
)

Arguments

professione

Character scalar identifying the profession to filter on.

data

A data.table with at least columns general_id, escoskill_level_3, and the column named by group_col.

group_col

Character name of the column used to identify professions. Defaults to "preferredLabel".

top_n

Integer maximum number of edges to retain (before largest-component filtering). Defaults to 30L.

min_weight

Integer minimum co-occurrence count. Edges below this threshold are dropped. Defaults to 10L.

Value

A data.table with columns from, to, weight. Returns an empty data.table (zero rows, same columns) when the profession has no data, a single skill, or no edges above min_weight.