Hello on Github! (and refurbishing TidyTuesday)

I have not posted in quite a while and this is mostly because I have been migrating my old website to this new one (which is also thanks to another new friend)! There are still some formatting kinks I want to work out and I need to figure out how to add both a comments section and an email subscriber button, but this will be my website for a long while to come (I am, for the first time, actually quite satisfied with it except for some very minor formatting issues).

We will be getting back into our regularly scheduled weekly posts now with a flavor towards stuff I’m learning or experiencing whilst working on NSF just to efficiently multitask (although that being said there are still a lot of post ideas I still have in store as well).

One other thing that has been in the back of my mind for a long while is TidyTuesday; I basically only did one post and abandoned it right after that. I realized right after the first post back then that it was not worth the time commitment required to make a quality post week after week. As I’ve been uploading all the blogs back onto the new website and looking at the poor state of TidyTuesday, I remembered a conversation I had with a friend and work team member. We have this running joke about whenever we encounter a roadblock, the solution is to just “run a regression” as the empirical economists do. One day, somehow the conversation diverged to this as well, and we were making jokes about running random regressions one random timeseries. This is what I think I will turn TidyTuesday into: a script that runs automatically every day that picks two random timeseries from the plethora of data APIs and runs a regression between the two. Then every week I pick one of the seven that I found interesting and talk a little bit more about the relationship between the two timeseries.

There are of course more details and functionalities I can add into the code to do this (such as picking some controls) but for now the project idea is just to do the automation and generate a systematic report each day (and then automatically push it to github). Regardless, I cannot promise much progress on this for a while between NSF and grad apps but do expect this to show its face by the end of the year.