israndom version 0.9.0 by Rudi Cilibrasi <cilibrar@cilibrar.com>

israndom is a randomness tester using nonclassical statistics from
information theory.  It applies ideas of Kolmogorov, Shannon, and
Cilibrasi in an easy-to-use parameter-free system.  Unlike traditional
randomness testers that often have complex and various parameters to
set, israndom uses only simple and intuitive parameters and works well
without any parameter adjustment at all in most cases.  By using recent
results from parameter-free machine learning research, it is able to
ask for less information from the user, and provides more accurate answers
in simpler form in most situations.  Therefore, israndom is preferred in
situations where it is not necessary to have an explicit statistical
model or understanding of the data (e.g. like the kind of insight a
scatter plot gives), and instead simply measure if data is random or not
in a binary sense.  israndom provides this information in an easy numerical
printed form for human consumption and as an exit return code for shell
scripts.  By using the expandable compression module feature of libcomplearn,
israndom is theoretically able to detect any kind of randomness
deficiency whatsoever.  Thus, it represents another way to look at
statistical analysis that simplifies use without sacrificing power.