Thank you for taking the time to make the video! Ive been wondering about copilot capabilities in R and found this. Did anyone find the copilot capabilities demonstrated here to be unimpressive? It essentially accomplished what an Rstudio shinyApp snippet easily does, but missed the libraries. Then it threw out some erroneous answers for what to use in place of runApp. Has anyone used copilot in non-contrived, contextually dependent, uncommon use cases? How does it perform? Does it perform better with particular languages than others? What does it do well? Ive heard the logic it generates is generally low quality and riddled with bugs and random dependency introductions. Does it do better just generating boilerplate? Can it do boilerplate embedded in context accurately?
In a few years we probably don't have to write code anymore. We just tell the algorithm to analize this, test that, produce an output and tweak it a little. Like in Star Trek, when they go "Computer, analize this and evaluate that". Absolutely mindblowing.
Thanks for the sharing. Just forgot the vscode is also a great tool besides the very prepared rstudio. Back to the execution problem you encountered. I guess that's bcs the shiny app is running and the vscode stops accepting any key stroke. Therefore, you should stop the app and the vscode shall return the control to you.
Hi everyone, it's now available in RStudio as well :) I've got a video explaining how to get setup with Copilot (+ChatGPT) in RStudio in just 5 minutes: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-t7NrkAeosog.html