A console app exploring C++ Standard Library's pseudorandom number generators (PRNG).

This project uses four PRNGs: mt19937, mt19937_64, rand_s, and rand. PRNGs based on the Mersenne Twister algorithm (mt19937, mt19937_64) are widely recommended for general-purpose use and include helpful distribution functions. rand_s is also good and easy to implement. While rand is quite old, produces poor results, and is not recommended for serious projects, I've included it to demonstrate the evolution of PRNGs in C++.

"Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin." -- John von Neumann

Von Neuman's reservations aside, you can do some pretty interesting things with PRNGs. This project attempts to demonstrate some of those interesting things and how the underlying functions work. Like my threadDemo, I've approached this project as a teaching tool for a hypothetical, informal crash-course on PRNGs in C++.

The app opens a console window on startup. Enter the command keywords separated by a space. Commands are case-sensitive. Function names are entered first, followed by any parameters. For example:

functionName parameter1 parameter2 ...

To generate ten pseudorandom integers, ranging from 0 to 10, using all four PRNGs, enter:

basic

Similarly, to generate an indefinite number of integers within a specific range, enter:

basic n min max

where n is the number of integers and min and max are lower and upper integer range limits, respectively. The basic function with user parameters is a threaded process and can be terminated with:

basic stop

Just for fun, generate a wall of text of 1M random characters by entering:

wall

A repeatable series of pseudorandom integers is also useful in some applications. A PRNG produces the same series of integers when initialized with the same seed integer. The following example generates ten such integers, again ranging from 0 to 10, with zero as the seed integer:

basicSeed

To generate an indefinite number of repeatable integers within a specific range, enter:

basicSeed n seed min max

where n is the number of integers, seed is the seed integer, and min and max are lower and upper integer range limits, respectively. The basicSeed function with user parameters is a threaded process and can be terminated with:

basicSeed stop

Flipping a coin is perhaps the simplest demonstration of true randomness. To simulate a coin flip, enter:

coinFlip

"Heads" or "Tails" will be printed to the console.

Rolling dice is a more complex manifestation of randomness. To simulate a dice roll, enter:

diceRoll numSides numDice

where numSides is the number of sides on the die and the number of dice is numDice. For example:

diceRoll 6 2

simulates the result of rolling two standard six-sided dice.

How many times would you need to flip a coin to get a streak of heads or tails? To find out, enter:

coinFlipStreak streak

where streak is the number of time heads or tails occurs in a row. coinFlipStreak is a threaded process and can be stopped by entering:

coinFlipStreak stop

Building on the previous example, how many dice rolls would be required to get the same side on all dice? Enter:

diceRollMatch numSides numDice

where numSides is the number of sides on the die and the number of dice is numDice. diceRollMatch is a threaded process and can be stopped by entering:

diceRollMatch stop

The 32-bit Mersenne Twister (mt19937) is the default PRNG for all chance-related functions. The default PRNG can be set to another by entering:

default set prngName

where prngName is rand, rand_s, mt32 for the 32-bit Mersenne Twister, or mt64 for the 64-bit Mersenne Twister.

To query the current PRNG, enter:

default query

To stop all active threads, enter:

stop all

To stop all active threads and exit the program, enter:

exit

randomDemo v1.1.0 [45 kB]

v1.1.0

• Added random class with mt19937, mt19937_64, rand_s, and rand PRNGs and related functions.
• Updated basic.h/cpp functions.
• Added chance.h/cpp for chance related PRNG functions.
• Updated README.

v1.0.0

• Initial commit.