This is an archive for the code to recreate the simulations
for the note:  A Compound Decision Bakeoff

The files sim[1234].R run the simulations for Tables 1-4.
The files sim[1234]a.R create the Tables 1-4.  The corresponding
oracle tables 5-8 are done with simor[1234].R and simor[1234]a.R.

The files ex1.R and oracle.R provide some preliminary testing.
The original Martin and Walker code available from Martin's website
was used for sim1.R, while a vectorized version given in ex1.R was
used in the remaining simulations since it was substantially faster.

Note that the simulations were run on a Mac pro using R's foreach()
construct and the random seeds for each instance of foreach are set
to be identical.  Roughly speaking, each of the 8 simulations requires
about 2 hours on my Mac Pro.

Roger Koenker
Oct 22 2013

total 152
-rw-r--r--  1 roger  roger  1922 Dec 14 10:53 Readme
-rw-r--r--  1 roger  roger  1475 Dec 14 10:50 ex1.R
-rw-r--r--  1 roger  roger   872 Dec 14 10:50 oracle.R
-rw-r--r--@ 1 roger  roger  1302 Dec 14 10:50 sim1.R
-rw-r--r--  1 roger  roger   438 Dec 14 10:50 sim1a.R
-rw-r--r--  1 roger  roger  1227 Dec 14 10:50 sim2.R
-rw-r--r--  1 roger  roger  1062 Dec 14 10:50 sim2a.R
-rw-r--r--  1 roger  roger  1288 Dec 14 10:50 sim3.R
-rw-r--r--  1 roger  roger   433 Dec 14 10:50 sim3a.R
-rw-r--r--  1 roger  roger  1216 Dec 14 10:50 sim4.R
-rw-r--r--  1 roger  roger   710 Dec 14 10:50 sim4a.R
-rw-r--r--  1 roger  roger   848 Dec 14 10:50 simor1.R
-rw-r--r--  1 roger  roger   444 Dec 14 10:50 simor1a.R
-rw-r--r--  1 roger  roger   810 Dec 14 10:50 simor2.R
-rw-r--r--  1 roger  roger   800 Dec 14 10:50 simor2a.R
-rw-r--r--  1 roger  roger   970 Dec 14 10:50 simor3.R
-rw-r--r--  1 roger  roger   439 Dec 14 10:50 simor3a.R
-rw-r--r--  1 roger  roger   796 Dec 14 10:50 simor4.R
-rw-r--r--  1 roger  roger   354 Dec 14 10:50 simor4a.R
