A Shiny interface to review potential duplicates
Source:R/manual_dedup_shiny.R
manual_dedup_shiny.Rddedup_citations() can return potential duplicates for manual review. This
function takes these potential duplicates and provides a Shiny interface to
review them and select those that should be deduplicated. The output can then
be passed to dedup_citations_add_manual() to complete the deduplication, or
be used to call this function again if manual review is not yet complete.
Usage
manual_dedup_shiny(df, cols = names(df))Arguments
- df
A dataframe containing potential duplicate entries, typically returned by
dedup_citations().- cols
A character vector of column names to display during the review process. By default, uses all columns in
df.
Value
The dataframe with an updated result column, indicating whether each
entry is a duplicate ("match") or not ("no_match"). This can be passed to
dedup_citations_add_manual() for completing the deduplication process.
The dataframe with a result column indicating whether the entry
constitutes a duplicate - to be passed to dedup_citations_add_manual()
Examples
if (FALSE) { # interactive()
# Perform deduplication
result <- dedup_citations(citations_df, keep_source="Embase")
# Manually review potential duplicates
manual_review <- manual_dedup_shiny(result$manual_dedup)
# Complete deduplication
final_result <- dedup_citations_add_manual(result$unique, additional_pairs = manual_review)
}