Generating Random Numbers

We have already used the Random class a little, but let's investigate this in more detail. The class Random enables you to create multiple random number generators that are independent of one another. Each object of the class is a separate random number generator. Any Random object can generate pseudo-random numbers of types int, long, float, or double. These numbers are created using an algorithm that takes a 'seed' and 'grows' a sequence of numbers from it. Initializing the algorithm twice with the same seed would produce the same sequence because the algorithm is deterministic.

The integer values generated will be uniformly distributed over the complete range for the type, and the floating point values will be uniformly distributed over the range 0.0 to 1.0 for both types. You can also generate numbers of type double with a Gaussian (or normal) distribution that has a mean of 0.0 and a standard deviation of 1.0. This is the typical bell-shaped curve that represents the probability distribution for many random events.

There are two constructors for a Random object. The default constructor will create an object that uses the current time from your computer clock as the seed value for generating pseudo-random numbers. The other constructor accepts an argument of type long that will be used as the seed.

Random lottery = new Random();          // Sequence not repeatable 
Random repeatable = new Random(997L);   // Repeatable sequence

If you use the default constructor, the sequence of numbers that is generated will be different each time a program is run, although beware of creating two generators in the same program with the default constructor. The time resolution used is one millisecond, so if you create two objects in successive statements they will usually generate the same sequence because the times used for the starting seed values will be identical. Random objects created using the same seed will always produce the same sequence, which can be very important when you are testing a program. Testing a program where the output is not repeatable can be a challenge! A major feature of random number generators created using a given seed in Java is that not only will they always produce the same sequence of pseudo-random numbers from a given seed, but they will also do so even on totally different computers.

import java.util.Random;
import java.io.IOException;

public class Dice {
  public static void main(String[] args) {
    System.out.println("You have six throws of a pair of dice.\n" + 
               "The objective is to get a double six. Here goes...\n");

    Random diceValues = new Random();      // Random number generator
    String[] theThrow = {"First ", "Second ", "Third ",
                         "Fourth ", "Fifth ", "Sixth "};
    int die1 = 0;                          // First die value
    int die2 = 0;                          // Second die value

    for(int i = 0; i < 6; i++)  {
      die1 = 1 + diceValues.nextInt(6);             // Number from 1 to 6
      die2 = 1 + diceValues.nextInt(6);             // Number from 1 to 6
      System.out.println(theThrow[i] + "throw: " + die1 + ", " + die2);

      if(die1 + die2 == 12) {                       // Is it double 6?
        System.out.println("    You win!!");        // Yes !!!
        return;
      }
    }
    System.out.println("Sorry, you lost...");
    return;
  }
}

You have six throws of a pair of dice.
The objective is to get a double six. Here goes...

First throw: 3, 2
Second throw: 1, 1
Third throw: 1, 2
Fourth throw: 5, 3
Fifth throw: 2, 2
Sixth throw: 6, 4
Sorry, you lost...

die1 = 1 + abs(diceValues.nextInt())%6;             // Number from 1 to 6