Wednesday, December 18, 2019

Ford vs. Ferrari

I recently was reminded to look again at Ignacio Sarmiento Barbieri's blog post
https://ignaciomsarmiento.github.io/2017/01/20/Quantile-and-OLS-Regression-with-Rmosek.html
that compares the speed of several implementations of quantile regression in Mosek with
that of the simplex implementation in my quantreg R package.  He finds that for problems
of modest size (n = 1000) the quantreg simplex implementation which is based on the 1972
algorithm of Barrodale and Roberts was substantial faster than any of the Mosek implementations.

However, in a more realistic econometric problem based on data from Josh Angrist (n = 97367, p = 5)
the simplex algorithm was quite a bit slower than the best of the Mosek implementations.  I wondered
how the interior point algorithms implemented in fortran in quantreg would perform.

Now, Mosek is clearly the Ferrari in this contest.  I've used Mosek in many different convex
optimization problems and always found it to be remarkably quick, responsive to special road
conditions and (especially) elegant in design.  So it requires a special sort of hubris to think
that the so-called Frisch-Newton implementations that were figuratively "made in my garage"
could compete, but lo and behold, here are the results:

Unit: milliseconds
                               expr       min        lq      mean    median
 qr.mosek.dual.rep(X, y, tau = tau)  346.9841  357.2982  365.9998  364.2814
        rq.fit.fnb(X, y, tau = tau)  193.7651  194.5455  202.1676  195.2470
       rq.fit.sfn(Xs, y, tau = tau)  330.3882  341.1280  365.0841  349.7058
         rq.fit.br(X, y, tau = tau) 7266.7237 7270.0682 7289.8437 7274.8538
        uq       max neval
  370.6428  462.1361   100
  206.7692  288.0764   100
  369.8432  451.3124   100
 7285.0870 7522.6136   100


The results for the basic (dense) Frisch-Newton appear as rq.fit.fnb while the sparse version of this
algorithm appear as rq.fit.sfn and the simplex version is rq.fit.br.

Sunday, October 27, 2019

Correlation

My daughter, the epidemiologist, carved this  pumpkin and sent the photo, which reminded my wife,
the Russian historian, that I "always get the flu when  a new Le Carre novel appears".  Correlation is not causation, it is only a Gaussian delusion, even Galton's regression to mediocrity figure doesn't quite exhibit a constant correlation.

Saturday, May 25, 2019

Theresa May needs a rest

I'm back from several wonderful days in Prague where the highlight was the new production of Love
for 3 Oranges -- part of the Prague Spring Music Festival.  Given yesterday's announcement by
Theresa May that she was stepping down as PM, I thought it would be appropriate to post a
cartoon, provided by Ivan Mizera, that sums up her regime very concisely in the proud tradition of Czech political humor.