-------------------------------------------------------------------------------
      name:  <unnamed>
       log:  randnumb_mt64.log
  log type:  text
 opened on:  30 Mar 2023, 16:01:59

. do randnumb_mt64

. * The DIEHARD tests of random number generators by George Marsaglia
. * can be found at 
. *
. * https://web.archive.org/web/20160125103112/http://stat.fsu.edu/pub/diehard/
. *
. * DIEHARD requires 3 million 32-bit integers generated by the random
. * number generator.
. 
. * Created for Stata 18 on 2023-03-30
. 
. clear all

. set obs 3000000
Number of observations (_N) was 0, now 3,000,000.

. 
. * reset the seed to the default
. set seed 123456789

. 
. * We use -long- storage type to get 32-bit integers
. *
. * In
. *
. *    McCullough, Brian D. 1999. "Assessing the Reliability of Statistical
. *    Software: Part II", The American Statistician 53(2): 149-159
. *
. * it indicates (on page 156) that the integers are to be calculated as
. *
. *    INT(-2147483649 + 4294967297*RAN)
. *
. gen long x = int(-2147483649 + 4294967297*runiform())

. 
. * The DIEHARD programs require a certain format for the data.  It wants
. * the numbers 10 on a line in hexadecimal representation in an ascii file.
. * There is then a DIEHARD utility program that takes this kind of file and
. * produces another binary file that DIEHARD actually uses.  McCullough
. * (see reference above) suggests just writing the numbers to an ascii file
. * and then writing a simple program to translate that into the format that
. * the DIEHARD utility program desires.
. 
. outfile x using randnumb_mt64.raw

. 
. * The C program raw2bin.c will take a file like randnumb.raw produced
. * above and create the binary file that the diehard program desires.
. 
. exit

end of do-file


. log close
      name:  <unnamed>
       log:  randnumb_mt64.log
  log type:  text
 closed on:  30 Mar 2023, 16:02:11
-------------------------------------------------------------------------------