Another newbie question from me.

I'm doing a really basic exercise to practice precedence rules in Java. The rules seem pretty straight forward to me, but something that I feel like should work simply doesn't.

The exercise is to write a console program that converts Farenheit temperature to Celsius, based on the following equation:


The class I wrote works perfectly and I understand how it conforms to precedence rules in Java:

Here's what I DON'T understand. Why does the following expression return a "0.0" regardless of what initial value I enter for Farenheit? I feel like it still conforms to Java's precedence rules, but obviously I'm still not getting something (and perhaps it has nothing to do with precendence).


I thought that the computations within the parenthesis were done first.

Sorry for the super basic questions! I'm doing this solo so I appreciate folks' taking the time out.

What you are not understanding is that when the compiler sees that statement, it assigns 5 and 9 to an integer. So, when you compute the integer division of 5/9, you get an integer back which is zero. Make it 5.0 and see what happens.

Ah that makes sense. So always express a number as x.0 if I want the compiler to treat it as a double.

Are there other ways to tell it to treat all values in an equation as doubles?

Thanks for setting me straight on that one!

You can use a cast on either operand (or both). I think that just getting into the habit of tacking on the ".0" onto any integer constant that you want to be treated as a double is more elegant (but that's just me).


Consistent with my personal style (and noting that you misspelled "Fahrenheit"), I probably would have written it

Cleaner appearing and not as many of those annoying parentheses to keep track of.


Bottom line:
If the division has only integers, the quotient will be the result of integer division.

If one operand of a binary arithmetic operation is a double and the other is an integer, the integer is "promoted" and the operation will be done with floating point arithmetic.

Regards,

Dave

That clears up a lot (down to the spelling).

So basically, because "5 / 9" is enclosed in parentheses, that computation is done first. And since both operands are integers, I end up with the integer 0.

And when I rewrite the expression as "5 * (F-32) / 9," the 5 is converted into 5.0 because (F-32) is treated as a double, and the 9 becomes 9.0 for the same reason. So that works, but of course doesn't look as nice.

"(5.0 / 9.0) * (fahrenheit - 32)" definitely wins on all fronts.

You've cleared up a lot for me. Thanks for taking the time out on the remedial end of things! I'm sure I'll be back in another chapter or two.