Take a data.frame of bibliographic information showing potential duplicates (as returned by find_duplicates), and return a data.frame of unique references.
extract_unique_references(x, matches)
Arguments
x: a data.frame to be subsetted
matches: either a vector of matches, e.g. as returned from find_duplicates, or a column name (specified as a number or a string) from x showing where matches are stored
Returns
a subsetted data.frame containing one entry for each group identified in matches.
Note
This function creates a simplified version of x, by extracting the reference from each group of 'identical' references that contains the most text. It is assumed that this is the most 'complete' record of those available in the dataset. This function does not merge data from multiple 'identical' records due to the potential for mis-matching that this approach would create.
See Also
find_duplicates for duplicate identification; screen_duplicates for an interactive alternative to duplicate removal.