Tidymodels have outcomes, clearly defined predict, easy to estimate performance
Tidyclust doesn't have outcomes, not clearly defined predict, no clear answer. #rstudioconf2022#rstudioconf#rstats
extract_cluster_aasignment to get ids of where things are placed
extract_centroids
predict (for some engines) #rstudioconf2022#rstudioconf#rstats
Metrics similar to yardstick.
tot_wss() total within sum of square errors
cluster_metric_set() to calculate multiple metrics (like yardstick::metric_set()). #rstudioconf2022#rstudioconf#rstats
Now that #RStudioConf2022 is over and I've had some time to process, it's time for a 🧵 to look back and summarize my copious notes. Hopefully this can be useful to someone other than just me! #RStudioConf#RStats#PyData
Moonshot goal: he decided to write and play a video game in #RStats. From what I've seen on twitter I suspect he also submitted it to #RStudioConf as a commitment device. #RStudioConf2022
Ha, yes, I've also been known to obsessively follow projects, Bryan! When I heard the tune 📦was coming but github.com/tidymodels/tune was a 404, I set up something to check it every couple hours then beepr::beep when it wasn't a 404 anymore 😁 #RStudioConf2022#RStudioConf#RStats
Keep the ui consistent. She's showing how we can have a rules based system so those of us who don't know how things SHOULD work can get it pretty much right. #RStudioConf2022#RStudioConf#RStats
What's it like to be a ui designer? It is awful. Also it is awesome. It just depends if things are working and/or pretty! #RStudioConf2022#RStudioConf#RStats
"how can we make shiny apps look better?" #noHTML#noCSS
* Look
* Functionality