Thread: Do { } Until Loop Continues Infinitly. Moronic Question (I Think)

Perhaps you mean to reset a after the inner loop exits? Also, use >= instead of !=, lesson 1.

> Please don't give me a 5 line code that does the same thing, that's not learning to me.

I hope that means you're going to do something about all those if statements; they're pretty ugly. You could at least use "else if"

Well, I won't give you the 5 line version, but 2 hints so you can figure out how to code the 5 line version yourself:

1. You can output a single character as well as a string:

cout << 'a';

2. You can do arithmetic on characters:

char ch = 'a' + 4; // Same as char ch = 'e'.

Oh, as a bonus tip, I recommend learning about for loops -- they usually are a better choice than do...while or while loops. For example, they avoid the common problem with nested loops that you had with not resetting a.

Originally Posted by iMalc

Any sensible beginner would learn an important lesson from this.

And a non-beginner might point out which part(s) as non-portable assumptions?

Incidentally, rand() is from <cstdlib>, and that header was not included. Also, the PRNG was not seeded with srand().

One way to condense the selection of a character is to use a string:

Code:

const char alphabet[] = "abcdefghijklmnopqrstuvwxyz";

// ...

std::cout << alphabet[rand() % 26];

Prelude won't be happy to see we're only using the low-order bits of rand()

Jasio, you can't just write some lines and expect it to work, adding cumbersome fixes along the way. Programming isn't just brute-forcing your way around and hoping it works.

Code:

#include <iostream>
#include <cstdlib>
#include <ctime>

int main()
{
    //get size of rand() output in bits
    size_t shift_rand_size;
    for(size_t i = RAND_MAX; i; i >>= 1, ++shift_rand_size);
    
    //shift it by 5 bits because 26 < 2^5
    if(shift_rand_size  >= 5) shift_rand_size -= 5;
    else shift_rand_size = 0;
    
    //seed rand()
    srand(time(NULL));
    
    //get user input
    size_t a, c;
    std::cout << "Attempts: ";
    std::cin >> a;
    std::cout << "Characters per attempt: ";
    std::cin >> c;
    
    //loop and print
    for(size_t i = 0; i < a; ++i)
    {
        for(size_t j = 0; j < c; ++j)
        {
            std::cout << char('a' + (rand() >> shift_rand_size) % 26);
        }
        std::cout << '\n';
    }
    return 0;
}

It would also do you good to learn the switch statement.

Sure, the higher bits are theoretically slightly more random.

No, it is not that the lower order bits are less random (though they may be, for a given poor implementation). Using modulo m for a range that is not evenly divisible by m leads to a (slight) change in the distribution.

But by throwing many lower order bits away you've increased the bias by many orders of magnitude. With your code, the letter 'a' is far far more likely to occur than 'z'.

I agree. I think something like this would be better:

Code:

int max = RAND_MAX - (RAND_MAX % 26) - 1;
int num; // random number generated
do {
	num = rand();
} while (num > max);
num %= 26; // change range to [0, 25] without altering distribution