This R package facilitates the reconciliation of multiple data sets containing plant names to each other via standardised World Flora Online name IDs. This enables the merging of data based on a global, consensus taxonomy in a reproducible way. It will also report on provenience of the name matching process and current taxonomy of the names used in the analysis for subsequent publication.
Think Reference Manager for plant names!
There is an existing WFO R package in CRAN for standardizing plant names (https://cran.r-project.org/web/packages/WorldFlora/index.html). It differs from wfor in that it requires a download of the complete name list in Darwin Core format whilst wfor relies on calling the API of the latest version of the list. The existing package can also use other lists of names in Darwin Core format to match against. The two packages are complementary. Hopefully we will see yet more packages released that work with WFO data in different ways.
One round of development has created this minimum viable product. It still needs more work but you are welcome to have a play. We plan to submit an improved version to CRAN by the end of 2023 In sh’Allah!
You can install the development version of wfor from GitHub with:
# install.packages("devtools")
devtools::install_github("rogerhyam/wfor")You have one or more data.frames containing observational or other data. One of the columns contains the names of plant species as strings of characters. They might be quite variable and come from a variety of sources. You need to standardize these strings so that they map to a shared lookup table. You will then be able to map between different data sets (merge data.frames) in your own study and those in other people’s studies in an explicit and repeatable way.
The function wfo_match_df_names takes your data.frame and returns a new version with four columns added:
- wfo_id contains the WFO ID of a name in the WFO Plant list.
- wfo_name contains the string of the full name as WFO has it.
- wfo_method contains a flag of how the name match was made, automatically or manually.
- wfo_path contains the taxonomic path to the name in the current WFO classification. This is a sanity check to be sure you haven’t matched your orchid name against a bryophyte by mistake!
You may have to run a data.frame through the function several times and do some manual checking to get 100% coverage. Once that is done you can keep a copy of the name cache which is a mapping between the strings found in your data and the WFO IDs. This can be used to speedup the processing of subsequent datasets and prevent you from making the same manual selection twice.
Coming soon: Once you have marked up your data with WFO IDs there are various functions that allow you to generate reports on your data for inclusion in publications. These might include species lists in alphabetical and/or taxonomic order with full details of place of publication, taxonomic sources and type specimens when available.
This is a basic example which shows you how to solve a common problem:
library(wfor)
# auto match the Belgian Magnoliopsida test data
`mags_example <- wfo_match_df_names(Belgian_Magnoliopsida_sp_ssp_var_2011, name_col="scientific_name", authors_col="authorship", interactive=FALSE)`
# check how many have been matched using the [wfo_stat_matches()] function.
`wfo_stat_matches(mags_example)`
# do a second pass interactively to resolve unmatched names (or until you get tired)
`mags_example <- wfo_match_df_names(mags_example, name_col="scientific_name", authors_col="authorship", interactive=TRUE)`
# do a third pass matching names to the nearest genus
`mags_example <- wfo_match_df_names(mags_example, name_col="scientific_name", authors_col= "authorship", fallback_to_genus=TRUE)`