Java's RNG

Lately I've been making small Java programs to refresh my skills.  I've been using C and Ruby this last semester for the more complicated programs (I'll post those soon), but I'd like to get to the point where I'm comfortable doing them in Java.

Today's program is a simple dice roll program that tests the proficiency of java.util.random.  I got the idea from /r/dailyprogrammer.

1:  /*  
2:  Name: Kevin Achenbach  
3:  Date: 7/12/2014  
4:  Project: test a RNG to see how random it actually is. Use 6 sided dice as the test  
5:  */  
6:  package dice;  
7:  import java.util.Random;  
8:  public class Dice {  
9:    static int diceRolls[] = {10,100,1000,10000,100000,1000000};  
10:    public static void main(String[] args) {  
11:      System.out.print("# of Rolls 1s   2s   3s   4s   5s   6s\n");  
12:      System.out.print("====================================================\n");  
13:      for(int i = 0; i< diceRolls.length; i++) {  
14:        System.out.format("%-11d", diceRolls[i]);  
15:        float percent[] = rollDice(6,diceRolls[i]);  
16:        //Display results   
17:        for(int j = 1; j<percent.length;j++) {  
18:          percent[j] = percent[j]*100;  
19:          String str = String.format("%2.02f", percent[j]);  
20:          System.out.print(str + "% ");  
21:        }  
22:        System.out.print("\n");  
23:      }  
24:    }  
25:    static float[] rollDice(int arraySize ,int numberOfRolls) {  
26:      //can assume the all array values are zero at initialization  
27:      int testData[] = new int[arraySize+1];  
28:      Random rand = new Random();  
29:      int x;  
30:      float percent[] = new float[arraySize+1];  
31:      for(int i = 0; i<numberOfRolls; i++) {  
32:        x = rand.nextInt(6)+1;  
33:        testData[x]++;  
34:      }  
35:      for(int j = 0; j<arraySize+1;j++) {  
36:        percent[j] = ((float)testData[j])/numberOfRolls;  
37:      }  
38:      return percent;  
39:    }  
40:  }  

Results:








The results were as expected.  Java.util.Random did not favor any number in particular as the number of rolls increased.  A few things I learned while doing this:

• line 36: converting two ints to a float by dividing is a bit tricky.  The conversion to a float happens after the division occurs.  This means that two integers dividing into a percentage will always come out as 0.0 unless the percentage is 100%.  To avoid this cast the numerator as a float before dividing.
• Java.util.Random is INCREDIBLY quick.  I ran the nextInt method over a million times yet the total run time was still under a second.