pow() function error

[nat45928]

Hi,

i have this sinppit of code to get all the individual digits in a number.

float tempNumber = 011010
for (short i = 6; i > 0; i--)
{
  cout << (static_cast<long>(tempNumber) % static_cast<long>(pow(10, i))) / static_cast<long>(pow(10, i-1)) << endl;
}

when i run this code i get an output of 0,1,1,0,2,0 why is it?

when i output just static_cast<long>(pow(10, i-1)) sometimes i get 99 instead of 100, why is this?

Nat

[TheIdeasMan]

An exercise for you:

Step through the code with a debbuger, with a watch list on your variables - see how they change line by line. Split the code into 3 lines, to make it easier to see what is going on. You have a % b / c, put a b c on their own lines.

Hint: It could be to do with floating point representation.

It's interesting tempNumber looks like a binary number (or is that coincidence)- does it have to be float?

[nat45928]

OK when i split it apart like so:

for (short i = 6; i > 0; i--)
	{
			long a = 11010;
			long b = static_cast<long>(pow(10, i));
			long c = static_cast<long>(pow(10, i-1));
			cout << a << endl << b << endl << c << endl;
			cout << a % b << endl;
			cout << (a % b)/c << endl << endl;
	}

Things work like a charm. I have no idea how to use a debugger, i have only used c++ for a few days now. I am a long time Java programmer though. you would be correct, i am writing a program to convert a base 2 decimal eg. 0.011010 into base ten eg. 0.42 the user inputs floats and that is why it is in a float type, i multiplied the inputted decimal in previous code so that i could do the % operator and get each individual digit.

[Peter87]

Not all floating point values can be stored exactly so rounding errors happens easily. If you want something exact you shouldn't use floating point numbers.

A number literal starting with 0 is an octal number. 011010 is the same as 4616.

[Disch]

Good God, man. Don't use pow() for this.

I'm not entirely sure what it is you're doing... but....

// -- BAD --
for( int i = 6; i > 0; --i )
{
  int a = 11010;
  int b = a / pow(10,i);
  cout << b;
}

// -- BETTER --
for(int i = 1000000; i > 0; i /= 10)
{
  int a = 11010;
  int b = a/i;
  cout << b;
}

EDIT:

As for extracting individual digits from a number (is that what you're doing?):

int num = 11010;
for(int digit = 1000000; digit > 0; digit /= 10)
{
  cout << ((num / digit) % 10);
}

[TheIdeasMan]

Peter87 wrote:

A number literal starting with 0 is an octal number. 011010 is the same as 4616.

Good Work Peter !! Eagle Eye. I am guessing a lot of people would never have seen that.

nat45928 wrote:

I have no idea how to use a debugger, i have only used c++ for a few days now. I am a long time Java programmer though.

Surely the process is similar, no matter which language you use. Maybe you are compiling / make from the cmd line? If that is so is a bit more tricky. I am sure you can find some tutorials on the web. There is a gdb one on this site.