• HPI Chorus

    Have you ever wondered what housing price indices would sound like if they were converted to music? Planet Money already answered that. But what if that music was worse? In this post, I’ll look at the result of converting state-level Housing Price Indices into music for each of the 50 states. Also, I’ll combine the HPI by region to get a sort of HPI chorus.

    Tags:    |   |   | 
  • McFar Points

    A few weeks back, a friend asked me if I was familiar with the concept of the “McFarthest Spot.” I wasn’t familiar with the idea, but a quick search informed me that it is the point in the continental United States where you are further away from a McDonald’s than at any other point. The concept seems to have first been considered by Stephen Von Worley as discussed in this post, and tends to randomly resurface on the internet from time to time, especially when the McFarthest Spot changes.

    Tags:    | 
  • Covid Commute

    The global pandemic has severely reduced peoples’ mobility world-wide, as regional and national lock-downs and stay-at-home orders have gone into effect. In this post, I’ll animate the impact that covid-19 has had on my daily commute/movements using my Google Location History data. Because I’m lucky enough to have a job that enabled me to work from home, and because my employer was proactive in putting us on work-from-home, I haven’t really been moving around much since mid March. As a result, the animation will get pretty boring around that point. 🙂

    Tags:    | 
  • Groundhog Day T-Testing

    While looking for some details on t-tests, I ran across this post from The Minitab Blog, which investigated Punxsutawney Phil’s powers of prognostication. The authors of that post tested whether there was a statistically significant difference in the average monthly temperature in the months immediately following Phil’s predictions when he saw his shadow vs when he didn’t. Detecting an increase in mean temperature when Phil saw doesn’t see his shadow would support Phil’s prognostication claims.

    Tags:    |   | 
  • Poisson Spam Riddler

    In this post, I’ll walk through a solution to FiveThirtyEight’s Riddler Classic problem from last week. In this problem, we are asked to consider a spam comment generation process. In this process, spam replies will respond to an initial post and to each subsequently generated spam reply. Spam will accumulate on each post/reply according to identical independent Poisson processes with a common rate of one comment per day (refer to the original prompt for a more clear explanation). We are asked: how many spam posts can we expect to have after three days?

    Tags:    |