/*************************** RANDOMC.H ***************** 2007-09-22 Agner Fog *
*
* This file contains class declarations and other definitions for the C++ 
* library of uniform random number generators.
*
* Overview of classes:
* ====================
*
* class CRandomMersenne:
* Random number generator of type Mersenne twister.
* Source file mersenne.cpp
*
* class CRandomMother:
* Random number generator of type Mother-of-All (Multiply with carry).
* Source file mother.cpp
*
*
* Member functions (methods):
* ===========================
*
* All these classes have identical member functions:
*
* Constructor(uint32 seed):
* The seed can be any integer. Usually the time is used as seed.
* Executing a program twice with the same seed will give the same sequence of
* random numbers. A different seed will give a different sequence.
*
* void RandomInit(uint32 seed);
* Re-initializes the random number generator with a new seed.
*
* void RandomInitByArray(uint32 seeds[], int length);
* In CRandomMersenne only: Use this function if you want to initialize with
* a seed with more than 32 bits. All bits in the seeds[] array will influence
* the sequence of random numbers generated. length is the number of entries
* in the seeds[] array.
*
* double Random();
* Gives a floating point random number in the interval 0 <= x < 1.
* The resolution is 32 bits in CRandomMother and CRandomMersenne.
*
* int IRandom(int min, int max);
* Gives an integer random number in the interval min <= x <= max.
* (max-min < MAXINT).
* The precision is 2^-32 (defined as the difference in frequency between 
* possible output values). The frequencies are exact if max-min+1 is a
* power of 2.
*
* int IRandomX(int min, int max);
* Same as IRandom, but exact. In CRandomMersenne only.
* The frequencies of all output values are exactly the same for an 
* infinitely long sequence. (Only relevant for extremely long sequences).
*
* uint32 BRandom();
* Gives 32 random bits. 
*
*
* Example:
* ========
* The file EX-RAN.CPP contains an example of how to generate random numbers.
*
*
* Optimized version:
* ==================
* Faster versions of these random number generators are provided as function
* libraries in asmlib.zip. These function libraries are coded in assembly
* language and support only x86 platforms, including 32-bit and 64-bit
* Windows, Linux, BSD, Mac OS-X (Intel based). Use asmlibran.h from asmlib.zip
*
*
* Non-uniform random number generators:
* =====================================
* Random number generators with various non-uniform distributions are available
* in stocc.zip (www.agner.org/random).
*
*
* Further documentation:
* ======================
* The file randomc.htm contains further documentation on these random number
* generators.
*
*
* Copyright:
============
*