Collapse the occurrences in utterances into groups — collapse

This function collapses all occurrences into groups sharing the same identifier, by default the stanzaId identifier ([[sid=..]]).

collapse_occurrences(
  parsedSource,
  collapseBy = "stanzaId",
  columns = NULL,
  logical = FALSE
)

Arguments

parsedSource: The parsed sources as provided by parse_source().
collapseBy: The column in the sourceDf (in the parsedSource object) to collapse by (i.e. the column specifying the groups to collapse).
columns: The columns to collapse; if unspecified (i.e. NULL), all codes stored in the code object in the codings object in the parsedSource object are taken (i.e. all used codes in the parsedSource object).
logical: Whether to return the counts of the occurrences (FALSE) or simply whether any code occurreded in the group at all (TRUE).

Value

A dataframe with one row for each value of of collapseBy and columns for collapseBy and each of the columns, with in the cells the counts (if logical is FALSE) or TRUE or FALSE (if logical is TRUE).

Examples

### Get path to example source
exampleFile <-
  system.file("extdata", "example-1.rock", package="rock");

### Parse example source
parsedExample <-
  rock::parse_source(exampleFile);
#> Warning: For graphs, attribute fontcolor had multiple values specified: 'Firebrick' & 'Black'; taking the first one.
#> Warning: For nodes of types matching regular expression 'grandchildCode2', duplicate attributes were found

### Collapse logically, using a code (either occurring or not):
collapsedExample <-
  rock::collapse_occurrences(parsedExample,
                             collapseBy = 'childCode1');

### Show result: only two rows left after collapsing,
### because 'childCode1' is either 0 or 1:
collapsedExample;
#>   childCode1 childCode1 childCode2 childCode3 childCode4 childCode5
#> 1          0          0          0          1          1          1
#> 2          1          2          1          0          0          0
#>   grandchildCode1 grandchildCode2 grandchildCode3 grandchildCode4
#> 1               1               3               1               1
#> 2               0               0               0               0
#>   grandchildCode5 grandchildCode6 grandchildCode7 someOtherCode
#> 1               1               1               1             1
#> 2               0               0               0             0

### Collapse using weights (i.e. count codes in each segment):
collapsedExample <-
  rock::collapse_occurrences(parsedExample,
                             collapseBy = 'childCode1',
                             logical=FALSE);