I just wanted to comment on my writing status.
I left my MMT manuscript at rest. I think I’ll do another reading pass in the next week, and if I’m happy, send it to my editor. It’s easier to spot problems in a text you’ve written after not reading it after a while, otherwise you tend to read what you wanted to write (and not what you actually read).
I wanted to write the r * estimation algorithms, in order to give a qualitative explanation of what the algorithms do. However, I was somewhat surprised by the behavior. I finally realize that the Kalman filter algorithm is not behaving the way I expected.
The Kalman filter is an ancient theory of control. It was no longer used in industrial applications that interested me, so I never paid much attention to it. From what I have seen of the control literature, its robustness has long been a concern. Sadly, it looks like I’m going down the rabbit hole. Given its importance in traditional empirical work, this is not trivial.
(c) Brian Romanchuk 2020