Prof. Manny Parzen, a pioneer of modern statistics, passed away in February, aged 87. I should have commented back then, but it’s still worth saying something today. I happened to be thinking of him this morning.

I did not know Manny personally. This makes it odd that I refer to him by his first name, but I do so in the same spirit that people who are fans of Bernie Sanders but don’t know him personally, including myself, refer to him as Bernie. This right away tells you something about Manny.

It also says something about Manny that he stayed active long after retirement age — and that he was actually using R! That is remarkable for a theoretician of his era, and indeed maybe even for the present era.

I discovered this at the 43rd Symposium on the Interface of
Computing Science and Statistics in 2012, where there was a special session honoring him. A couple of his former PhD students spoke, including my old professor, Don Ylvisaker, and Manny himself gave a talk.

Manny was pretty frail at the time, using a walker, yet mentally sharp as ever, and as I mentioned, still quite active in research. In fact, the young man sitting next to me turned out to be a current PhD student working under Manny.

At any rate, what jolted me was that in the midst of his talk, on the subtle theoretical properties of histograms and the like, he suddenly started talking about R! Specifically, he brought in R’s quantile() function, and went into a comparison of the 9 different options quantile() offers the user for defining quantiles. This really impressed me, and illustrates how much R has become a central tool.

By the way, after the talk I went up and introduced myself, and told him a joke that he had told many years ago when he gave a talk in the UC Davis Stat Dept. I won’t waste bandwidth here by going into the details of the joke, but it concerned students who guess randomly on exams. He laughed heartily at his own joke, and said, “It’s still true!”