In any project, the data we work with represents more than just facts and figures. It represents people, their places, their cultures — unique assemblages of environments and the communities that rely on them. And the way they generate and record data is just as varied and nuanced! So how can we make comparisons?
For those of us working in the research or data realm, we know all too well the many different ways people may label and organize data, even those in the same lab or working group. Here at OHI, we are fortunate to have data that encompasses people groups and ecosystems from all over the world, collected from a multitude of global data sources. This complexity and diversity can also amplify a particularly irksome challenge that we often face in synthesizing data sets: matching identifiers and naming conventions.
In this post, I’m going to showcase just a couple of the issues that I’ve encountered as an OHI Fellow, as well as functions and R packages that came to my rescue! Hopefully, this information can help you out too, if you ever find yourself in a frenzy of missing variables or unmatched data frames.
As you can imagine, variation in the spelling of country names is derived from a wide variety of factors. Most prominently, researchers tend to record labels and identifiers in their native languages, and compilations of data translated from one language to another will differ based on the type of translation.
For example, in an English translation:
South Korea
vs The Republic of Korea / 대한민국)To address these inconsistencies, people often choose a single standard of naming conventions and convert any mismatches. This has been especially helpful in my projects at OHI, and creates reproducible workflows that other fellows can also use.
Rather than manually renaming countries or creating an extremely long
case_when() list, you can use the countrycode package!
countrycode() to the Rescuecountrycode() from the countrycode
package
allows you to convert country names into UN naming conventions in
several different languages, including dialects.
According to its GitHub page, the package can “convert to and from 600+ variants of country names in different languages and formats.”
It also supports:
At OHI, we use this to convert a mix of region codes (like UN georegion IDs) into a unified naming standard — making dataset joining much easier.
```{r}
# install.packages("countrycode")
library(countrycode)
# Example: Convert common names to ISO3C codes
countrycode(c("South Korea", "Germany", "Brazil"),
origin = "country.name",
destination = "iso3c")
```
For many of the OHI goals— such as:
their score calculations involve marine organism data, typically structured as observational records with species names and identifiers. Like country names, these species labels often vary in format and spelling.
If some organisms are recorded by species name while others are only labeled by genus or a variation of the common name, it becomes challenging to accurately analyze patterns in catch, population size, or geographic distribution.
For example:
Thunnus maccoyii-57
Thunnus spp.
Pacific Bluefin Tuna
Bluefin Tuna
likely refer to the same or similar organisms, but inconsistent naming makes them difficult to analyze or compare.
Sometimes you may not be able to get more specific information about a species (e.g. inability to downscale broad categories, limitations in metadata or data sharing agreements), but you can convert names to a standardized framework, which allows for comparison.
rfishbase for Marine SpeciesThe rfishbase
package
standardizes taxonomic and common names based on two updating databases:
FishBase and SeaLifeBase.
According to its GitHub page
rfishbase's“simplified access protocol relies onduckdbfsfor direct reads of tables”.
This package is incredibly helpful for
Converting scientific names to common names (sci_to_common())
Retrieving multiple common names across languages (common_names())
Validating names to catch typos or outdated taxonomy (validate_names())
These functions helped me standardize species data and avoid issues like double-counting or missing species entirely due to taxonomic changes.
```{r}
# install.packages("rfishbase")
library(rfishbase)
# Get all common names for a species
common_names("Thunnus maccoyii")
# Convert scientific to common name
sci_to_common("Thunnus maccoyii")
# Validate names
validate_names(c("Thunnus maccoyii", "Thunnus atlanticus", "Scomber
japonicus"))
```
I hope these packages and tips help you as much as they’ve helped me during my time as OHI Fellow!
Whether you’re working with global region codes or marine biodiversity data, there’s likely an R package that can make your workflow more efficient, more reproducible, and less frustrating.
Happy coding! ✨