.f float suffix question

[slouchcouchgamer]

I'm trying to get my head around this piece of code, which demonstrates the use of integer literal suffix:

	{ // Demonstrating .f
		float a = 3.2;
		if ( a == 3.2 )
			std::cout << "a is equal to 3.2" << std::endl;
		else
			std::cout << "a is not equal to 3.2" << std::endl;

		float b = 3.2f;
		if ( b == 3.2f )
			std::cout << "b is equal to 3.2f" << std::endl;
		else
			std::cout << "b is not equal to 3.2f" << std::endl;
	}

The first if statement is false, whereas the second if statement where .f is used is true.

How is that possible? If a was an int, I understand that the decimal place would be lost, and a would equal 3. But a float can hold decimal values, and is equal to 3.2...no?

[Ganado]

a float variable holds less decimal places than a double variable. The number "3.2" cannot be represented exactly in binary to a computer, similar to how 1/7 can't be represented exactly in base-10.

3.2 is initially a double. 3.2f is initially a float.
float a= 3.2; casts down from a double to a float when assigned, so you lose precision.

3.2f might be represented as something like:
11.00110011001100₂
3.2 (a double) might be represented as:
11.00110011001100110011001100110011₂

*Note I didn't actually count the number of decimal places, just note that double usually has about double of float.

These numbers are not exactly equal, so the check is false.

[slouchcouchgamer]

Thanks Ganado.

I wasn't relating it to binary, makes much more sense now.

I still find it tricky to know exactly when to use .f, I presume it's pretty much only when you are using relational operators with floats.