Programming Forums > Application Development > C

hipolito · Jul 6th, 2006, 8:00 AM

I think even if I do this:

number float;
number = stof (printf("%.2f\n", number));

not sure it stof is string to float, but anyway I would still have all digits with zeros right?

Can I just set it to 2 digits lik 44.23? maybe I have to change it to short or whatever!

thx for the help guys.

Ooble · Jul 6th, 2006, 8:25 AM

You can use the round, roundf and roundl functions to round numbers to the nearest integer. Given that, we can use simple multiplication and division to round to x decimal places.

(Toggle Plain Text)

float number[3] = { 89.27218, 0, 0 };
number[1] = round(number[0]); // 89
number[2] = round(number[0] * 100) / 100); // 89.27

frankish · Jul 6th, 2006, 11:49 AM

Another way to do it would be to mask out all but the first 2 digits in the decimal part of the number. This would require knowledge of how a float is stored on your machine.

Narue · Jul 6th, 2006, 2:40 PM

Quote:

But as he says he needs to use it in the program not as a output, is it not possible at the beginning when you make the my_float to also define how many number my_float is supposed to store? I.e my_float=[4] or something?

No. You can do it for some values, but not others. You can do it for constant values, but nothing that changes over time. If you need to use the value for some kind of comparison, set a fudge factor to the appropriate radix and realize that floating-point is a bitch to work with.

(Toggle Plain Text)

#include <stdio.h>
#include <math.h>

int main ( void )
{
  double a = 87.99876;
  double b = 87.98876;
  const double fudge = .01;

  if ( fabs ( a - b ) <= fudge )
    puts ( "Identical with fudge" );
  else
    puts ( "Different" );

  a = 87.99876;
  b = 87.99765;

  if ( fabs ( a - b ) <= fudge )
    puts ( "Identical with fudge" );
  else
    puts ( "Different" );

  return 0;
}

DaWei · Jul 6th, 2006, 4:36 PM

There are actual quantities and then there are representations of quantities and the failings thereof. Grab your hand calculator and compare 1/3*3 with 1*3/3. Everyone knows they're the same (except your calculator). Another kind of problem is that floating point methods store a much larger range of numbers in the same space as integers with a much smaller range. A second's thought will raise suspicions that something had to go, in the tradeoff. Your machine likes to scale numbers using powers of 2, whereas most of us like to scale numbers using powers of 10. Things won't always work out cleanly.

frankish · Jul 6th, 2006, 5:24 PM

Quote:

Originally Posted by DaWei

There are actual quantities and then there are representations of quantities and the failings thereof. Grab your hand calculator and compare 1/3*3 with 1*3/3. Everyone knows they're the same (except your calculator). Another kind of problem is that floating point methods store a much larger range of numbers in the same space as integers with a much smaller range. A second's thought will raise suspicions that something had to go, in the tradeoff. Your machine likes to scale numbers using powers of 2, whereas most of us like to scale numbers using powers of 10. Things won't always work out cleanly.

Interesting...

Game_Ender · Jul 6th, 2006, 7:48 PM

Quote:

Originally Posted by DaWei

Grab your hand calculator and compare 1/3*3 with 1*3/3.

You sir, need a newer calculator

.

*Whips out TI-84*
1/3*3 = 1
1*3/3 = 1

It works even if you store 1/3 in a variable then multiply by 3. Gotta love the TI-84, but your point is still true, its just not the best example.

DaWei · Jul 6th, 2006, 7:54 PM

Quote:

Grab your xxxxxxxxxxxxxx ten-cent hand calculator...

:o :p

grumpy · Jul 10th, 2006, 8:53 AM

Quote:

Originally Posted by Game_Ender

You sir, need a newer calculator

.

*Whips out TI-84*
1/3*3 = 1
1*3/3 = 1

It works even if you store 1/3 in a variable then multiply by 3. Gotta love the TI-84, but your point is still true, its just not the best example.

I'm not going out to buy a TI-84, but I would guess that's because your calculator stores numbers internally to a higher precision than it displays them. What happens if you compute 1/3*3-1?

hipolito · Jul 17th, 2006, 9:14 AM

Well, this all started couse I can mak a number like 83.389999
become a 83.39 (as a string), but when I put it back to float it becomes
83.390002, and that is all I dont want it.

I guess I can not use float, what is the kind I should use for 83.39?

Jul 6th, 2006, 8:25 AM	#12
Ooble I eat cake for breakfast. Join Date: Jul 2004 Location: In my box. Posts: 4,434 Rep Power: 9	You can use the round, roundf and roundl functions to round numbers to the nearest integer. Given that, we can use simple multiplication and division to round to x decimal places. (Toggle Plain Text) float number[3] = { 89.27218, 0, 0 }; number[1] = round(number[0]); // 89 number[2] = round(number[0] * 100) / 100); // 89.27 float number[3] = { 89.27218, 0, 0 }; number[1] = round(number[0]); // 89 number[2] = round(number[0] * 100) / 100); // 89.27 __________________ Me :: You :: Them

Jul 6th, 2006, 4:36 PM	#15
DaWei Resident Grouch Join Date: Jun 2005 Posts: 6,453 Rep Power: 10	There are actual quantities and then there are representations of quantities and the failings thereof. Grab your hand calculator and compare 1/33 with 13/3. Everyone knows they're the same (except your calculator). Another kind of problem is that floating point methods store a much larger range of numbers in the same space as integers with a much smaller range. A second's thought will raise suspicions that something had to go, in the tradeoff. Your machine likes to scale numbers using powers of 2, whereas most of us like to scale numbers using powers of 10. Things won't always work out cleanly. __________________ Abstraction doesn't make it impossible to write bad code; it makes it possible to write superior code. Contributor's Corner: Grumpy on C++ Exceptions DaWei on Pointers

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Jul 6th, 2006, 8:00 AM	#11
hipolito Programmer Join Date: Feb 2005 Posts: 41 Rep Power: 0	I think even if I do this: number float; number = stof (printf("%.2f\n", number)); not sure it stof is string to float, but anyway I would still have all digits with zeros right? Can I just set it to 2 digits lik 44.23? maybe I have to change it to short or whatever! thx for the help guys.

Jul 6th, 2006, 11:49 AM	#13
frankish Hobbyist Programmer Join Date: Oct 2005 Location: Ohio Posts: 177 Rep Power: 0	Another way to do it would be to mask out all but the first 2 digits in the decimal part of the number. This would require knowledge of how a float is stored on your machine.

Jul 17th, 2006, 9:14 AM	#20
hipolito Programmer Join Date: Feb 2005 Posts: 41 Rep Power: 0	Well, this all started couse I can mak a number like 83.389999 become a 83.39 (as a string), but when I put it back to float it becomes 83.390002, and that is all I dont want it. I guess I can not use float, what is the kind I should use for 83.39?

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Programming Forums > Application Development > C

set a number to 2 digits only