another simple C question - Tilted Forum Project Discussion Community

Dilbert1234567 · 02-07-2004, 12:35 PM

Ok this program is the same as the last one, i am to enter 3 variables height width and length and it spits out the volume, the square footage and the ratio between the length and the width, it does everything except the ratio it rounds to the nearest hole number, even though it is stored as a double.

Thanks in advance.

Quote:

#include <stdio.h>

main(void)
{
int nRoomLength=0, nRoomWidth=0, nRoomHeight=0;

double dWidthToLength=0 ; /* width/length */

short int nRoomVolume=0 ; /* nRoomWidth * nRoomLength * nRoomHeight*/

unsigned int nRoomArea=0 ; /* nRoomWidth * nRoomLength */

printf("Please enter a value for the room width:");
scanf("%d", &nRoomLength);
printf("Please enter a value for the room length::");
scanf("%d", &nRoomWidth);
printf("Please enter a value for the room height::");
scanf("%d", &nRoomHeight);

nRoomArea = nRoomLength * nRoomWidth;
nRoomVolume = nRoomLength * nRoomWidth * nRoomHeight;
dWidthToLength = (nRoomWidth / nRoomLength);
printf("The area of the floor of the room is: %d\n", nRoomArea);

printf("The volume of the room is: %d\n", nRoomVolume);

printf("The ratio of the width to the length is: %lf\n\n", dWidthToLength);

printf("Variable Dump: \n\n");

printf("Variable\tNumber of\tAddress of\tValue in\n");
printf("Name\t\tBytes\t\tVariable\tVariable\n");

printf("nRoomWidth\t%d\t\t%p\t%d\n",sizeof(nRoomWidth),&nRoomWidth,nRoomWidth);
printf("nRoomWidth\t%d\t\t%p\t%d\n",sizeof(nRoomLength),&nRoomLength,nRoomLength);
printf("nRoomWidth\t%d\t\t%p\t%d\n",sizeof(nRoomHeight),&nRoomHeight,nRoomHeight);
printf("nRoomWidth\t%d\t\t%p\t%lf\n",sizeof(dWidthToLength),&dWidthToLength,dWidthToLength);
printf("nRoomWidth\t%d\t\t%p\t%d\n",sizeof(nRoomVolume),&nRoomVolume,nRoomVolume);
printf("nRoomWidth\t%d\t\t%p\t%d\n",sizeof(nRoomArea),&nRoomArea,nRoomArea);

}

Silvy · 02-07-2004, 03:11 PM

I'm no C - programmer but maybe the operands for the division need to be doubles for a result to return a double?

dValueResult = dValueWidth / dValueLength
(or perhaps only one of the two)

kel · 02-07-2004, 03:33 PM

Your not really describing what your problem is.

I am going to let you in on a little secret, real programmers don't bother to save every teeny tiny bit here and there.

If you designing an object that might have millions of instantiation... well then I might understand if you used shorts when you don't need the capacity.

Mixing data types is generally a pain in C/Java like languages because there are strict rules that govern how the compiler will do things. You end up sending explicit casting instructions (which waste time) and don't end up with the hugest performance or space gains. And casting makes your code ugly... really ugly.

In an object with 10 members if you make one of them a short, you only save a few bytes. Even with millions of them your only saving some constant multiple of space. It might sound important but in the big scheme of things these aren't the make or break performance determinants. The exception being realtime/interactive applications.

In this case I would say just use floats throughout. If this were Java I would say doubles.

Quadraton · 02-07-2004, 03:51 PM

Quote:

Originally posted by Silvy
I'm no C - programmer but maybe the operands for the division need to be doubles for a result to return a double?

dValueResult = dValueWidth / dValueLength
(or perhaps only one of the two)

Yes, that's exactly it. It's a common mistake made when doing division in C (one that I've made myself many times

)

Let's say you have something like:

Code:

int    nDenominator  = 2;
int    nNumerator   = 3;
double dResult;
dResult = nNumerator / nDenominator;

You would expect the value of dResult to be 1.5, except that it's not. Instead, it ends up holding 1.0. Why? Because in order for floating point division to work, either one or both the numerator, and denominator in the equation need to be a double (at least in Win32 it does, I didn't have a chance to verify this with any other compiler).

You can fix this code in either one of two ways.

1) Change the variables nNumerator or nDenominator to be a double
2) Typecast either one or both of the variables to be a double (i.e. (double) nNumerator / (double) nDenominator ).

In either case, you end up getting the answer 1.5, which is exactly what you should be getting.

Quadraton · 02-07-2004, 04:02 PM

Quote:

Originally posted by kel
Your not really describing what your problem is.

I am going to let you in on a little secret, real programmers don't bother to save every teeny tiny bit here and there.

If you designing an object that might have millions of instantiation... well then I might understand if you used shorts when you don't need the capacity.

Mixing data types is generally a pain in C/Java like languages because there are strict rules that govern how the compiler will do things. You end up sending explicit casting instructions (which waste time) and don't end up with the hugest performance or space gains. And casting makes your code ugly... really ugly.

In an object with 10 members if you make one of them a short, you only save a few bytes. Even with millions of them your only saving some constant multiple of space. It might sound important but in the big scheme of things these aren't the make or break performance determinants. The exception being realtime/interactive applications.

In this case I would say just use floats throughout. If this were Java I would say doubles.

I'm sorry, but huh??

I can see exactly where you're coming from, but you shouldn't cloud a simple issue with big talk like instanstiation and memory management. It's unnecessary, and just adds chaos to confusion.

kel · 02-07-2004, 04:03 PM

He is having trouble because he is trying to use smaller data types to save space. I am explaining to him why this isn't always a good idea.

Quadraton · 02-07-2004, 04:22 PM

Quote:

Originally posted by kel
He is having trouble because he is trying to use smaller data types to save space. I am explaining to him why this isn't always a good idea.

I don't think he's trying to do anything other that write a simple program. I doubt he's very concerned with optimizing his memory usage at this point.

Not to put any words into Dilbert1234567's mouth, or anything.

kel · 02-07-2004, 04:25 PM

Neither should I :-)
Dilbert why did you mix ints,shorts, doubles, and unsigned ints?

KnifeMissile · 02-07-2004, 10:18 PM

Yeah, I think Dilbert1234567 is confused about the use of data types. However, I disagree with kel as to why he's confused. I simply think that Dilbert1234567 has no understanding of the roles of the different variable types.

I mean, who would use integers to represent area or volume?

Generally speaking, integers are used for counting or indexing, and floats (or doubles) are used for measurements...

cheerios · 02-08-2004, 12:29 AM

*sigh* men. ok, long and short of it: an int, divided by an int, will always return an int. you need to use a more precise type of variable to store your data. or cast your ints as doubles before your devison. for the "why" sa well as the how, see above.

Dilbert1234567 · 02-08-2004, 12:31 AM

i have to use the difrent memory sizes for the home work, to see waht and how they function, my problem is that the program is rounding dWidthToLength when it should not be rounding it off.

cheerios · 02-08-2004, 02:49 AM

dWidthToLength = (nRoomWidth / nRoomLength);

ok. dWidthToLenght is a double. but nRoomWidth and nRoomLenght are INTEGERS. you need to open your book up, and look up Integer division. it's not the same as normal division. if you divide integers, you will get an integer answer, that is equal to the floor of the decimal division of the 2 values. so 2/4 = 2. 2/5 will ALSO =2. ALWAYS. because you're deviding INTEGERS.

to prevent this, you can do 2 things.
1: cast nRoomWidth and nRoomLength as doubles or floats.
2: make nRoomWidth and nRoomLength doubles or floats.

that's it. those are your options.

Dilbert1234567 · 02-09-2004, 10:51 PM

ya i just found that so i taked on *1.0 and that fixes it, int*int * float = float

02-07-2004, 03:11 PM	#2 (permalink)
Silvy paranoid Location: The Netherlands	I'm no C - programmer but maybe the operands for the division need to be doubles for a result to return a double? dValueResult = dValueWidth / dValueLength (or perhaps only one of the two) __________________ "Do not kill. Do not rape. Do not steal. These are principles which every man of every faith can embrace. " - Murphy MacManus (Boondock Saints)

02-07-2004, 03:33 PM	#3 (permalink)
kel WARNING: FLAMMABLE Location: Ask Acetylene	Your not really describing what your problem is. I am going to let you in on a little secret, real programmers don't bother to save every teeny tiny bit here and there. If you designing an object that might have millions of instantiation... well then I might understand if you used shorts when you don't need the capacity. Mixing data types is generally a pain in C/Java like languages because there are strict rules that govern how the compiler will do things. You end up sending explicit casting instructions (which waste time) and don't end up with the hugest performance or space gains. And casting makes your code ugly... really ugly. In an object with 10 members if you make one of them a short, you only save a few bytes. Even with millions of them your only saving some constant multiple of space. It might sound important but in the big scheme of things these aren't the make or break performance determinants. The exception being realtime/interactive applications. In this case I would say just use floats throughout. If this were Java I would say doubles. __________________ "It better be funny"

02-07-2004, 04:03 PM	#6 (permalink)
kel WARNING: FLAMMABLE Location: Ask Acetylene	He is having trouble because he is trying to use smaller data types to save space. I am explaining to him why this isn't always a good idea. __________________ "It better be funny"

02-07-2004, 04:25 PM	#8 (permalink)
kel WARNING: FLAMMABLE Location: Ask Acetylene	Neither should I :-) Dilbert why did you mix ints,shorts, doubles, and unsigned ints? __________________ "It better be funny"

02-08-2004, 12:31 AM	#11 (permalink)
Dilbert1234567 Devils Cabana Boy Location: Central Coast CA	i have to use the difrent memory sizes for the home work, to see waht and how they function, my problem is that the program is rounding dWidthToLength when it should not be rounding it off. __________________ Donate Blood! "Love is not finding the perfect person, but learning to see an imperfect person perfectly." -Sam Keen

02-07-2004, 10:18 PM	#9 (permalink)
KnifeMissile Location: Waterloo, Ontario	Yeah, I think Dilbert1234567 is confused about the use of data types. However, I disagree with kel as to why he's confused. I simply think that Dilbert1234567 has no understanding of the roles of the different variable types. I mean, who would use integers to represent area or volume? Generally speaking, integers are used for counting or indexing, and floats (or doubles) are used for measurements...

02-08-2004, 12:29 AM	#10 (permalink)
cheerios Banned Location: 'bout 2 feet from my iMac	sigh men. ok, long and short of it: an int, divided by an int, will always return an int. you need to use a more precise type of variable to store your data. or cast your ints as doubles before your devison. for the "why" sa well as the how, see above.

02-08-2004, 02:49 AM	#12 (permalink)
cheerios Banned Location: 'bout 2 feet from my iMac	dWidthToLength = (nRoomWidth / nRoomLength); ok. dWidthToLenght is a double. but nRoomWidth and nRoomLenght are INTEGERS. you need to open your book up, and look up Integer division. it's not the same as normal division. if you divide integers, you will get an integer answer, that is equal to the floor of the decimal division of the 2 values. so 2/4 = 2. 2/5 will ALSO =2. ALWAYS. because you're deviding INTEGERS. to prevent this, you can do 2 things. 1: cast nRoomWidth and nRoomLength as doubles or floats. 2: make nRoomWidth and nRoomLength doubles or floats. that's it. those are your options.

02-09-2004, 10:51 PM	#13 (permalink)
Dilbert1234567 Devils Cabana Boy Location: Central Coast CA	ya i just found that so i taked on 1.0 and that fixes it, intint * float = float __________________ Donate Blood! "Love is not finding the perfect person, but learning to see an imperfect person perfectly." -Sam Keen