Count the number of times each pair
of items appear together within a group.
For example, this could count the number of times two words appear within documents.
This function has referred to pairwise_count in widyr package,
but with very different defaults on several parameters.
pairwise_count_dt(
.data,
.group,
.value,
upper = FALSE,
diag = FALSE,
sort = TRUE
)A data.frame.
Column name of counting group.
Item to count pairs, will end up in V1 and V2 columns.
When FALSE(Default), duplicated combinations would be removed.
Whether to include diagonal (V1==V2) in output. Default uses FALSE.
Whether to sort rows by counts. Default uses TRUE.
A data.table with 3 columns (named as "V1","V2" and "n"), containing combinations in "V1" and "V2", and counts in "n".
dat <- data.table(group = rep(1:5, each = 2),
letter = c("a", "b",
"a", "c",
"a", "c",
"b", "e",
"b", "f"))
pairwise_count_dt(dat,group,letter)
#> V1 V2 n
#> <char> <char> <int>
#> 1: a c 2
#> 2: a b 1
#> 3: b e 1
#> 4: b f 1
pairwise_count_dt(dat,group,letter,sort = FALSE)
#> V1 V2 n
#> <char> <char> <int>
#> 1: a b 1
#> 2: a c 2
#> 3: b e 1
#> 4: b f 1
pairwise_count_dt(dat,group,letter,upper = TRUE)
#> V1 V2 n
#> <char> <char> <int>
#> 1: a c 2
#> 2: c a 2
#> 3: a b 1
#> 4: b a 1
#> 5: b e 1
#> 6: e b 1
#> 7: b f 1
#> 8: f b 1
pairwise_count_dt(dat,group,letter,diag = TRUE)
#> V1 V2 n
#> <char> <char> <int>
#> 1: a a 3
#> 2: b b 3
#> 3: a c 2
#> 4: c c 2
#> 5: a b 1
#> 6: b e 1
#> 7: e e 1
#> 8: b f 1
#> 9: f f 1
pairwise_count_dt(dat,group,letter,diag = TRUE,upper = TRUE)
#> V1 V2 n
#> <char> <char> <int>
#> 1: a a 3
#> 2: b b 3
#> 3: a c 2
#> 4: c a 2
#> 5: c c 2
#> 6: a b 1
#> 7: b a 1
#> 8: b e 1
#> 9: e b 1
#> 10: e e 1
#> 11: b f 1
#> 12: f b 1
#> 13: f f 1
# The column name could be specified using character.
pairwise_count_dt(dat,"group","letter")
#> V1 V2 n
#> <char> <char> <int>
#> 1: a c 2
#> 2: a b 1
#> 3: b e 1
#> 4: b f 1