Thread by @WeAreRLadies

messy, realistic datasets are **extremely** useful for learning, teaching, exploring, and practicing your #rstats skills. so where do you find datasets like this?!

here are some practical tips you can use to find messy datasets! 🧵

you may be familiar with the #TidyTuesday project as a fantastic community effort to practice data visualization.

BUT did you know it's an incredible resource for finding datasets on topics ranging from pollution to board games to tech adoption?

github.com/rfordatascienc…

did you know some of the weekly #TidyTuesday dataset contributions actually include cleaning scripts? that means you can practice either on the original/raw data OR the slightly-transformed ready-to-visualize data.

Take a look at SF Rent data for example: github.com/rfordatascienc…

#2. if you are an undergraduate or graduate student, your institution likely has a data librarian! you can reach out and ask them about datasets in your topics of choice!

when i worked as a data librarian in the social sciences, I LOVED when a student asked for help finding data

#3. open data portals! you can likely find governmental data portals wherever you are in the world at some geographic level! in the US, i often gravitate to cities (because i find them fascinating and complex). take a look at food inspection data in NYC!

data.cityofnewyork.us/Health/DOHMH-N…

#4. site-limited searching on google. this is honestly a general-use superpower for finding what you need online use the "site:" parameter like "climate data .csv site:github.com".

I particularly like trawling through github in this way to find unexpected data

and re: github, are you familiar with Awesome Lists? here is the awesomelist of Public Datasets on github - truly a fantastic place to do some digging: github.com/awesomedata/aw…

one more note on #TidyTuesday - if you want some inspiration about how to make decisions about data cleaning for these datasets, there's a whole world of TidyTuesday videos/streams!

here's a very sweet example from the talented @lisalendway

https://twitter.com/lisalendway/status/1507174162992926726

more messy dataset ideas! here are two great suggestions from @alittlestats

* The Data is Plural newsletter - data-is-plural.com

"a weekly newsletter of useful/curious datasets, published by @jsvine "

* The {mosaicData} package: github.com/ProjectMOSAIC/…

"Project MOSAIC is a community of educators working to develop a new way to introduce mathematics, statistics, computation and modeling to students in colleges and universities."

thanks @alittlestats !