diehard_bitstream.h

Dieharder是一个是一个测试一序列伪随机数的随机性能的测试套件

/*
 * diehard_bitstream test header.
 */

/*
 * function prototype
 */
void diehard_bitstream(Test **test,int irun);

static Dtest diehard_bitstream_dtest = {
  "Diehard Bitstream Test",
  "\n\
#==================================================================\n\
#                  Diehard Bitstream Test.\n\
# The file under test is viewed as a stream of bits. Call them  \n\
# b1,b2,... .  Consider an alphabet with two \"letters\", 0 and 1 \n\
# and think of the stream of bits as a succession of 20-letter  \n\
# \"words\", overlapping.  Thus the first word is b1b2...b20, the \n\
# second is b2b3...b21, and so on.  The bitstream test counts   \n\
# the number of missing 20-letter (20-bit) words in a string of \n\
# 2^21 overlapping 20-letter words.  There are 2^20 possible 20 \n\
# letter words.  For a truly random string of 2^21+19 bits, the \n\
# number of missing words j should be (very close to) normally  \n\
# distributed with mean 141,909 and sigma 428.  Thus            \n\
#  (j-141909)/428 should be a standard normal variate (z score) \n\
# that leads to a uniform [0,1) p value.  The test is repeated  \n\
# twenty times.                                                 \n\
# \n\
# Note that of course we do not \"restart file\", when using gsl \n\
# generators, we just crank out the next random number. \n\
# We also do not bother to overlap the words.  rands are cheap. \n\
# Finally, we repeat the test (usually) more than twenty time.\n\
#\n\
# WARNING!  Many RNGs that \"should\" pass this test marginally\n\
# fail or are weak (and can be pushed to failure with increasing\n\
# numbers of psamples).  This suggests either an implementation bug\n\
# or an error in the presumed target data.  The tests should not\n\
# be used to perform an assessment of RNGs until this issue is\n\
# resolved.   rgb\n\
#==================================================================\n",
  100,
  2097152,
  1
};

/*
 * Global variables (none in this case)
 */