update: the data in the Jupyter notebook have been updated with the results from the Championship event in Boulder CO. The content below has not yet been updated.
I took my daugher to the Strider Bike race in Charlotte. It seemed like most of the participants were boys. I wanted to take a look at the names in the race results posted to the Strider Bikes website to see if I could verify my hunch.
https://www.striderbikes.com/charlotte (accessed July 2018 -- use Wayback Machine to find the page then).
conda create --name stridergender python=3.6
activate stridergender
conda install jupyter pandas pip
pip install gender_guesser
python StriderGender.py
deactivate| Categorization | Number |
|---|---|
| male | 35 |
| unknown | 10 |
| female | 4 |
| mostly_male | 3 |
Nearly all of the male names were very clearly male. Of the unknown names, 2 or 3 might have been female. The mostly_male names could honestly go either way. Another gender detector package might do a better job than this one (let me know if you have suggestions!). It seems like a package that used the whole name (including surname) and supported more ethnicities would be more accurate.
Limitations: these results only included the top 16 finishers in the race classes that had more participants than 16.
Another note, if you are a parent of one of these kids and want me to replace the last name with "Smith" or something for privacy reasons, let me know at david.bradway@gmail.com