[JAVA] Division always results in 0?

[FakezZ]

So here's the code:

for (int i = 0; i < numOfRuns; i++)		{			a = (double) i;			x += i;			percentage = (i / numOfRuns) * 100;			System.out.println(percentage + "% completed");		}

How can the percentage always be 0? numOfRuns in this scenario is 900000, if it matters...

[Si3Rra_7]

Congratz u just divided by 0 . Jk i don't know anything about programming

[FakezZ]

Congratz u just divided by 0 . Jk i don't know anything about programming

Lel... If I did, it would just throw an exception . And I was like: AN ANSWER? SO FAST? WUT????

[Jade]

Congratz u just divided by 0 . Jk i don't know anything about programming

Then don't post.

 

Use % instead of / -  / is "divides evenly", % is divides into.

[Si3Rra_7]

Then don't post.

 

Use % instead of / -  / is "divides evenly", % is divides into.

Don't be anti-joke chicken , if we don't have fun why do we live ?

[Guest]

percentage = ((float)i/numOfRuns)*100;

[FakezZ]

Then don't post.

 

Use % instead of / -  / is "divides evenly", % is divides into.

Tried it, it just gives random numbers

[Grrizz]

Tried it, it just gives random numbers

/ is divide, % is remainder/modulus, your problem is i / numOfRuns is truncating so will always be equal to 0 as it will always be a fraction, AleksaNS solution will fix it.

[ManWithBeard1990]

Try (i * 100)/ numOfRuns

instead of (i / numOfRuns) * 100

 

That prevents the rounding error that you have from occuring.

You see, (i / numOfRuns) is less than 1 and java rounds this to 0 before multiplying by 100.

 

EDIT: Changed around one of the parentheses to better illustrate what I meant but you don't really need them.

[FakezZ]

percentage = ((float)i/numOfRuns)*100;

and it will still not give you percentage.

I solved it by writing ((float) i / (float) numOfRuns), cause it would give crazy big numbers otherwise. Anyway, it works now, thanks!

[madknight3]

Use % instead of / -  / is "divides evenly", % is divides into.

Modulus (%) is remainder. http://www.cafeaulait.org/course/week2/15.html

 

If i and numOfRuns in this code are integers, then you're doing integer division which gets rid of the decimal part.

percentage = (i / numOfRuns) * 100;

One of them needs to be a float/double. You could cast with (float) or (double) like @AleksaNS suggested, or just use the a variable you're setting for some reason.

percentage = (a / numOfRuns) * 100;// orpercentage = ((float) i / numOfRuns) * 100;percentage = ((double) i / numOfRuns) * 100;

[Jade]

 

Modulus (%) is remainder. http://www.cafeaulait.org/course/week2/15.html

Aha, didn't entirely expect to be correct. Thanks for correcting me.

[ManWithBeard1990]

I solved it by writing ((float) i / (float) numOfRuns), cause it would give crazy big numbers otherwise. Anyway, it works now, thanks!

That works too but I advise you to try my way too, if you can. It's good programming practice to multiply before you divide when you do both, to prevent rounding errors.

[FakezZ]

 

Modulus (%) is remainder. http://www.cafeaulait.org/course/week2/15.html

 

If i and numOfRuns in this code are integers, then you're doing integer division which gets rid of the decimal part.

percentage = (i / numOfRuns) * 100;

One of them needs to be a float/double. You could cast with (float) or (double) like @AleksaNS suggested, or just use the a variable you're setting for some reason.

percentage = (a / numOfRuns) * 100;// orpercentage = ((float) i / numOfRuns) * 100;percentage = ((double) i / numOfRuns) * 100;

Yeah I tried quite a few things, but they didn't seem to work (possibly some mistake on my part). Now it works perfectly

[madknight3]

That works too but I advise you to try my way too, if you can. It's good programming practice to multiply before you divide when you do both, to prevent rounding errors.

That still doesn't work if it's integer division. Still need to work with float/double

[ManWithBeard1990]

That still doesn't work if it's integer division. Still need to work with float/double

It's only to 1% accurate but it does work (and for things that don't take too long that's really all the user really wants to know). Just multiply before you divide. Trust me, I've programmed before.

[madknight3]

It's only to 1% accurate but it does work (and for things that don't take too long that's really all the user really wants to know). Just multiply before you divide. Trust me, I've programmed before.

Sure it's valid if you want to truncate the decimal place and not round up when required. Depends on the use and accuracy needed.

[ManWithBeard1990]

Sure it's valid if you want to truncate the decimal place and not round up when required.

It's not entirely accurate I know, but does it really need to be? For some applications it probably does, but doing it without casting is computationally less expensive. Your way is entirely correct, I know. More correct than mine. I almost proposed the same solution as you did but decided not to because I wanted to show OP (@FakezZ) that when performing integer calculations it's better to multiply first and then divide instead of doing it the other way round. For floats this does not matter this much, but for ints it just does.

[madknight3]

It's not entirely accurate I know, but does it really need to be? For some applications it probably does, but doing it without casting is computationally less expensive. Your way is entirely correct, I know. More correct than mine. I almost proposed the same solution as you did but decided not to because I wanted to show OP (@FakezZ) that when performing integer calculations it's better to multiply first and then divide instead of doing it the other way round. For floats this does not matter this much, but for ints it just does.

I agree