Finished initial text on RandomStream

Author: nmcl

docs/manual.adoc

@@ -383,7 +383,7 @@ public  abstract  class  RandomStream
     protected  RandomStream  ();
     protected  RandomStream  (long  MGSeed,  long  LCGSeed);
 
-    protected final double Uniform (); 
+    protected final double uniform (); 
 }
 ----
 
@@ -394,3 +394,11 @@ Y[i+1] = Y[i] * 55  mod 226
 THIS NEEDS FIXING!!
 
 , where the period is 224, and the initial seed must be odd (Thanks to Professor I. Mitrani for his help in developing this.)
+
+The uniform() method uses the linear congruential generator (seed is LCGSeed, with the default value of 1878892440L) based on the algorithm in link:http://algs4.cs.princeton.edu/home/[this text], and the results of this are shuffled with the multiplicative generator (see is MGSeed, with a default value of 772531L) as suggested by link:http://www.amazon.com/Art-Computer-Programming-Volume-Seminumerical/dp/0201896842[Maclaren and Marsaglia], to obtain a sufficiently uniform random distribution, which is then returned.
+
+The error() method returns a chi-square error measure on the uniform distribution function.
+
+By abstract method getNumber must be provided by derived classes, and is used to obtain a uniform means of accessing random numbers.
+
+The  RandomStream class  returns  a  large  sequence  of  random  numbers,  whose  period  is 2^24. However, unless the seeds are modified when each random distribution class is created, the starting position in this sequence will always be the same, i.e., the same sequence of numbers will be obtained. To prevent this, each class derived from RandomStream has an additional parameter for one of its constructors which indicates the offset in this sequence from which to begin sampling.