Grouping comparer function

[renkun-ken]

list.group() uses identical() to group list elements by a field while list.class() uses contains(), i.e. match(x, table, nomatch = 0L) > 0L to classify list elements by a field.

The consequence is that list.group() is by default extremely strict. For example,

> library(rlist)
> x <- list(1,1,1L,1L)
> list.group(x, .)
$`1`
$`1`[[1]]
[1] 1

$`1`[[2]]
[1] 1


$`1`
$`1`[[1]]
[1] 1

$`1`[[2]]
[1] 1


> list.class(x, .)
$`1`
$`1`[[1]]
[1] 1

$`1`[[2]]
[1] 1

$`1`[[3]]
[1] 1

$`1`[[4]]
[1] 1

This output may be ambiguous and sometimes is undesired.

[renkun-ken]

The groups are determined by unique() which differentiates 1 and 1L.

> unique(list(1,1L))
[[1]]
[1] 1

[[2]]
[1] 1