Combine documents in a dfm by a grouping variable

Source: R/dfm_group.R

Combine documents in a dfm by a grouping variable, which can also be one of the docvars attached to the dfm. This is identical in functionality to using the "groups" argument in dfm().

dfm_group(x, groups = NULL, fill = FALSE, force = FALSE)

Arguments

x

a dfm
groups

either: a character vector containing the names of document variables to be used for grouping; or a factor or object that can be coerced into a factor equal in length or rows to the number of documents. NA values of the grouping value are dropped. See groups for details.
fill

logical; if TRUE and groups is a factor, then use all levels of the factor when forming the new "documents" of the grouped dfm. This will result in documents with zero feature counts for levels not observed. Has no effect if the groups variable(s) are not factors.
force

logical; if TRUE, group by summing existing counts, even if the dfm has been weighted. This can result in invalid sums, such as adding log counts (when a dfm has been weighted by "logcount" for instance using dfm_weight()). Does not apply to the term weight schemes "count" and "prop".

Value

dfm_group returns a dfm whose documents are equal to the unique group combinations, and whose cell values are the sums of the previous values summed by group. Document-level variables that have no variation within groups are saved in docvars. Document-level variables that are lists are dropped from grouping, even when these exhibit no variation within groups.

Setting the fill = TRUE offers a way to "pad" a dfm with document groups that may not have been observed, but for which an empty document is needed, for various reasons. If groups is a factor of dates, for instance, then using fill = TRUE ensures that the new documents will consist of one row of the dfm per date, regardless of whether any documents previously existed with that date.

Examples

corp <- corpus(c("a a b", "a b c c", "a c d d", "a c c d"),
               docvars = data.frame(grp = c("grp1", "grp1", "grp2", "grp2")))
dfmat <- dfm(corp)
dfm_group(dfmat, groups = "grp")
#> Document-feature matrix of: 2 documents, 4 features (25.0% sparse) and 1 docvar.
#>       features
#> docs   a b c d
#>   grp1 3 2 2 0
#>   grp2 2 0 3 3
dfm_group(dfmat, groups = c(1, 1, 2, 2))
#> Document-feature matrix of: 2 documents, 4 features (25.0% sparse) and 1 docvar.
#>     features
#> docs a b c d
#>    1 3 2 2 0
#>    2 2 0 3 3

# equivalent
dfm(dfmat, groups = "grp")
#> Document-feature matrix of: 2 documents, 4 features (25.0% sparse) and 1 docvar.
#>       features
#> docs   a b c d
#>   grp1 3 2 2 0
#>   grp2 2 0 3 3
dfm(dfmat, groups = c(1, 1, 2, 2))
#> Document-feature matrix of: 2 documents, 4 features (25.0% sparse) and 1 docvar.
#>     features
#> docs a b c d
#>    1 3 2 2 0
#>    2 2 0 3 3