Help!! trying to get 6 random numbers

when i compile it, i get the 6 random numbers upto 49. after few tries, sometimes I get the same numbers !! so how do i fix this?

Code:

---

#include <stdio.h>
#include <math.h>
#include <time.h>
#include <stdlib.h>
int main (void)
{
int a, b, c, d, e, f;
// print title
printf("*** 6/49 LOTTERY PICKER ***\n");


// seed generator with time of day
srand(time(NULL));
a= rand()%49+1;
b= rand()%49+1;
c= rand()%49+1;
d= rand()%49+1;
e= rand()%49+1;
f= rand()%49+1;


printf("My Lucky Numbers Are %d, %d, %d, %d, %d, %d\n", a, b, c, d, e, f);


return 0;

---

What do you mean by "same numbers"? Do you sometimes get the exact same list of numbers, for a-f? Or do you get a single number that is repeated in the set of lottery numbers?

If you get the the exact same numbers for all six lottery numbers, consider this: time(NULL) only gives you time to the nearest second. Thus if you run your program twice in the same second, srand() gets the same seed value and thus repeats the sequence. Remember, it only produces pseudo-random numbers.

If you get a single number (or two perhaps) repeated, then you must realize that is the nature of a random (or pseudo-random) number generator. It may repeat a number from time to time. Sometimes the space between a number and a repeat of that number is quite small, only a few numbers in between. This problem is made worse by the fact that you use modulo 49 (% 49), since it takes the ~4 billion possible values and squishes them all into one of 49 boxes. That really increases the chance of a repeat. There is a solution to this, which is pretty easy.

Simply store the numbers for your lottery in an array of ints. Then, shuffle that array (Fisher), and pick the first six numbers. A small example, picking 3 of 6:

Code:

---

[1, 2, 3, 4, 5, 6]
// shuffle it
[3, 1, 6, 4, 5, 2]
// picking the first 3 gives
3, 1, 6

---

Rather than time(NULL), you can ask the standard library for the time in microseconds. This value is very unlikely to be the same between two invocations of your program.

Code:

---

struct timeval tp;
gettimeofday(&tp, NULL);
unsigned long micros = tp.tv_sec * 1000000;
micros += tp.tv_usec;
srand((unsigned)(micros % UINT_MAX));

---

This tutorial should help you

FAQ > Generate random numbers? - Cprogramming.com

I use this code for random numbers, and up until now it seems to work pretty well. :-)

Code:

---

#include <ctime>
#include <cstdlib>
#include <iostream>
using namespace std;

bool random = true;

int r( int r1 , int r2 )
{
    if( random )
    {
        srand( ( int ) time ( 0 ) );
        random = false;
    }

    int r3 = rand();
    return r3 = rand() % ( r2 - r1 + 1 ) + r1;
}

int main()
{
    cout << r( 1 , 40 ) << endl;
    cout << r( 1 , 40 ) << endl;
    cout << r( 1 , 40 ) << endl;

    cin.get();
    return 0;
}

---

I'm still a beginner at programming, though, so the code might be a bit crude.
I hope it helps you, though. :-)

Quote:

---

Originally Posted by Kain

I use this code for random numbers, and up until now it seems to work pretty well. :-)

Code:

---

#include <ctime>
#include <cstdlib>
#include <iostream>
using namespace std;

bool random = true;

int r( int r1 , int r2 )
{
    if( random )
    {
        srand( ( int ) time ( 0 ) );
        random = false;
    }

    int r3 = rand();
    return r3 = rand() % ( r2 - r1 + 1 ) + r1;
}

int main()
{
    cout << r( 1 , 40 ) << endl;
    cout << r( 1 , 40 ) << endl;
    cout << r( 1 , 40 ) << endl;

    cin.get();
    return 0;
}

---

I'm still a beginner at programming, though, so the code might be a bit crude.
I hope it helps you, though. :-)

---

That is C++ code

This is a C thread :P

Quote:

---

Originally Posted by c99tutorial

Rather than time(NULL), you can ask the standard library for the time in microseconds. This value is very unlikely to be the same between two invocations of your program.

---

That's unlikely to be the problem here. This is code for a "LOTTERY PICKER", which indicates that his problem is in wanting unique values rather than what he has at the moment which will obviously produce duplicates on occasion.

blackendstars, I put the question to you, what ways can you think of to ensure that there are no duplicates? What would you do if you were getting the numbers by rolling 49-sided dice (if such a thing existed) and wanted them to be unique?

Code:

---

#include <stdio.h>
#include <time.h>

int main()
{
    srand((unsigned)time(NULL));
    int numbers,loop;

    for(loop=1;loop<=6;loop++)
{
    numbers=rand()%49+1;
    printf("Random #%d = %d\n",loop,numbers);

}
return(0);
}

---

If you want N unique random numbers, a simple way is the following

1. Make an empty list capable of holding N items.
2. Generate a random number x.
3. If x does not exist in L, then: add x to L.
4. Go to 2.

Break the loop after you have added N items to L. For example, you can use a while loop with a counter that increments each time you successfully find a new random number. For this strategy to work, the range of valid random numbers must be large compared to N. If N gets too large, this strategy has a high probability of getting "stuck" in the loop.

Quote:

---

Originally Posted by c99tutorial

If N gets too large, this strategy has a high probability of getting "stuck" in the loop.

---

In which case you would be better off generating the entire range, then shuffling to obtain the N numbers.

You guys are making it seem way too complicated.

Quote:

---

Originally Posted by LTA85

You guys are making it seem way too complicated.

---

I don't think you understand the problem.

Your program in post #8 is a simple example of finding six random numbers within a certain range.

However, the OPs requirements seem to include making sure there are no duplicate numbers. That is a little bit more involved.

Note that this was already mentioned by iMalc in the post above yours (post #7).

If you want to use the shuffle method, you can do it simply using qsort and a compar function that returns -1 or 1 randomly

Code:

---

int cointoss_compar(const void *a, const void *b) {
    (void)a; // ignore these
    (void)b;
    return rand() < (RAND_MAX/2) ? -1 : 1;
}

void shuffle(int *arr, int n) {
    qsort(arr, n, sizeof(*arr), cointoss_compar);
}

int main(void) {
    int numbers[10] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10};
    shuffle(numbers, 10);
    print(numbers, 10);
    return 0;
}

---

Quote:

---

Originally Posted by c99tutorial

If you want to use the shuffle method, you can do it simply using qsort and a compar function that returns -1 or 1 randomly

---

No point though since implementing a basic version Fisher-Yates/Knuth shuffle is not that hard, especially with the help of online resources.

Quote:

---

Originally Posted by c99tutorial

If you want to use the shuffle method, you can do it simply using qsort and a compar function that returns -1 or 1 randomly

---

Actually, that's rather dangerous. It's implementation dependent of course, but I know that the partition technique I've used at times avoids having multiple conditions in the innermost loops because it assumes that you can't go through the entire array without hitting an element that is not greater than the splitter. (I.e. you when it runs into the splitter itself). Violating this assumption could cause it to run off the end of the buffer.

Even more importantly, make sure you don't do this in C++ using std::sort. You risk not only causing the above, but also violate the strict-weak-ordering assumption that std::sort has, which it may well pull you up on in a debug build.

I found the following in the standard, which confirms the above is a hack. I thought it would be fine... it definitely won't work with std::sort. But even with qsort the behaviour violates the requirement of a total ordering

ISO/IEC 9899:201x, 7.22.5

Quote:

---

When the same objects (consisting of size bytes, irrespective of their current positions
in the array) are passed more than once to the comparison function, the results shall be
consistent with one another. That is, for qsort they shall define a total ordering on the
array, and for bsearch the same object shall always compare the same way with the
key.

---