Large pseudo random numbers

[JosAH]

How about 2**160? It does involve with BigIntegers right. So I need to use nextLong() method....thank u again

Hold your horses for a second: I wrote that for numbers up to 2**64 you can use
longs; for larger numbers the nextLong() method is obviously of no use.

Have a look at the constructor BigInteger(Stri ng val, int radix) and think
of the nextBoolean() method in the Random class: i.e. generate each bit one
by one, stick them in a String and let the BigInteger constructor do the dirty job.

kind regards,

Jos

ps. this is as far as I go because the entire assignment is no fun anymore after
this last reply.

[shana07]

Hold your horses for a second: I wrote that for numbers up to 2**64 you can use
longs; for larger numbers the nextLong() method is obviously of no use.

Have a look at the constructor BigInteger(Stri ng val, int radix) and think
of the nextBoolean() method in the Random class: i.e. generate each bit one
by one, stick them in a String and let the BigInteger constructor do the dirty job.

kind regards,

Jos

ps. this is as far as I go because the entire assignment is no fun anymore after
this last reply.

I do really appreciate your effort to advise me....thank you very much.
I just don't know how to use nextBoolean() in here...Could you please check my code. Kindly please advise me again (please don't feel bad about my bad programming skill).......Th ank you

600458679029219 288571181967070 626676968651940 7577
926708444724260 528080317242268 129922715806415 7338
......//Keep writing 50 random numbers
...//I do need it to start from 0 (inclusive) to max 49 decimal digits (2**160).


Here is my program...

public class MyBigRandom
{    
  public static void main(String[] args) 
  {
    String out="";    
  
    BigInteger b = null;        
    BigInteger a = new BigInteger("49", 10);  
    
    for (int i = 0; i < 50; i++) 
    {              
        out = getRandom(a.intValue());        
        System.out.println(out);
    }     
  }

  public static final String getRandom(int n) 
  {
        if (n < 1) return "";
        StringBuffer sb = new StringBuffer(n);
        java.util.Random r = new java.util.Random();
        
      //  while(r.nextBoolean())
        for (int i = 0; i < n; i++) 
        {
            sb.append(r.nextInt(10));
        }
           return sb.toString();        
  }
}

[JosAH]

The Random.nextBool ean() method returns random true/false values. For every
true value your can append a '1' to your StringBuffer, for every false value you
add a '0' to your StringBuffer. Finally use the constructor BigInteger(buf. toString(), 2)
and voila.

kind regards,

Jos

[shana07]

The Random.nextBool ean() method returns random true/false values. For every
true value your can append a '1' to your StringBuffer, for every false value you
add a '0' to your StringBuffer. Finally use the constructor BigInteger(buf. toString(), 2)
and voila.

kind regards,

Jos

I have one question about this code, kindly please take a look ...

public static final String getRandom(int n) 
    {
        if (n < 1) return "";
        StringBuffer sb = new StringBuffer(n);
        java.util.Random r = new java.util.Random();

        while(r.nextBoolean())       
        {
            sb.append(r.nextInt(10));  // What is this for? as I think it will seed number from 0 - 9. Then where is the relation for max value 2^160?
        }        
        return sb.toString(); 
        //return(new BigInteger(sb.toString(), 2));
    }

[JosAH]

This is a complete give away, I know, but this thread is going nowhere. Run
the following code snippet and see for yourself:

Random r= new Random(System.currentTimeMillis());
int nofbits= 40;
...
StringBuffer sb= new StringBuffer();
		
for (int i= 0; i < nofbits; i++)
	sb.append(r.nextBoolean()?'1':'0');
		
System.out.println(sb);

I know you have to put it in a main() method and all, but just try this; I hope you
can figure out the rest for yourself after all the tips, hints and code snippets that
have been given to you.

kind regards,

Jos

[shana07]

This is a complete give away, I know, but this thread is going nowhere. Run
the following code snippet and see for yourself:

Random r= new Random(System.currentTimeMillis());
int nofbits= 40;
...
StringBuffer sb= new StringBuffer();
		
for (int i= 0; i < nofbits; i++)
	sb.append(r.nextBoolean()?'1':'0');
		
System.out.println(sb);

I know you have to put it in a main() method and all, but just try this; I hope you
can figure out the rest for yourself after all the tips, hints and code snippets that
have been given to you.

kind regards,

Jos

Thank you Jos.. I managed to get it by your help of course. God bless you.
One question from me...as I need the max length is 160 bits, so I changed nofbits = 160; Why all numbers exhibit about equal lengths? that's 49 or 48 decimal digits?

Thank you from me....

[JosAH]

Thank you Jos.. I managed to get it by your help of course. God bless you.
One question from me...as I need the max length is 160 bits, so I changed nofbits = 160; Why all numbers exhibit about equal lengths? that's 49 or 48 decimal digits?

Thank you from me....

There's a very small chance that the generated number will be:
000 ... <160 zero bits here> ... 000 and the decimal representation would be
a single zero. Think of it this way: how large would the percentage be for all
numbers between 0 and 100 to be less than 10 (one digit).

For all numbers between 0 and 10**49 only 1 percent will be less than 10**47

kind regards,

Jos

[shana07]

There's a very small chance that the generated number will be:
000 ... <160 zero bits here> ... 000 and the decimal representation would be
a single zero. Think of it this way: how large would the percentage be for all
numbers between 0 and 100 to be less than 10 (one digit).

For all numbers between 0 and 10**49 only 1 percent will be less than 10**47

kind regards,

Jos

Herm I got your point Jos.
One last favor, after all, do I need to try different approach to get uniformly distributed of lenghts I mean between those range? It just that in my mind to generate test cases that involve with very large number it must be varies of numbers and lenght, what do you think?
Again, kindly advise me...thank you

[JosAH]

Herm I got your point Jos.
One last favor, after all, do I need to try different approach to get uniformly distributed of lenghts I mean between those range? It just that in my mind to generate test cases that involve with very large number it must be varies of numbers and lenght, what do you think?
Again, kindly advise me...thank you

If you generate n bits at random (uniformly distributed as the Random.nextBool ean()
method does) you get uniformly distributed numbers in the range [0, 2**n) so
there's nothing to worry about. e.g. if you generate uniformly distributed numbers
in the range [0, 1000) just one percent of the number will be less then 10.

kind regards,

Jos

[shana07]

If you generate n bits at random (uniformly distributed as the Random.nextBool ean()
method does) you get uniformly distributed numbers in the range [0, 2**n) so
there's nothing to worry about. e.g. if you generate uniformly distributed numbers
in the range [0, 1000) just one percent of the number will be less then 10.

kind regards,

Jos

Ok Jos. Thanks for your valueable advice and of cause your effort to reply my questions. Feeling better now..
please let me be advised again in future...