C# math question.

[the86d]

I haven't coded since Borland C++ for DOS, and Quickbasic for DOS, but this might be what you are looking for to round?:
ANSWER1=INT(7/2+.5)
ANSWER2=INT(5/7+.5)

Add .5, then truncate as an INT.

Old-school rounding.

[njineermike]

It doesn't round; instead it drops the decimal portion of the number.

So yes, 5/2 = 2.5

(int)2.5 = 2

This is what I always thought was correct.

[Deedle]

I haven't coded since Borland C++ for DOS, and Quickbasic for DOS, but this might be what you are looking for to round?:
ANSWER1=INT(7/2+.5)
ANSWER2=INT(5/7+.5)

Add .5, then truncate as an INT.

Old-school rounding.

Might not give the result you're expecting in all cases, sadly.

INT(-3.9 + 0.5) ...... gives 3, not 4

There are many types of rounding, so it's crucial to know what is expected at the design/requirements level. A short and incomplete list:

Round away from zero
Round toward zero
Round to nearest even (AKA bankers rounding)
Round to nearest odd
Round up
Round down
....

[bob7122]

Thank you all for the help, I really appreciate it.

[HUTCH 7.62]

[smashycrashy]

If you don't mind a bit of free advice from someone who flunked out of school, then became an honors student, a teacher, and is now writing (vhdl) code for a satellite project I can't talk about in public:

if you wonder "what will happen with this code", write the code and run it to find out.

"Why does this strange thing happen" is a much better question than "what will happen". Your job, basically, for the rest of your career, is to impress your (teachers/bosses/coworkers/lovers) with how hard working, curious, and resourceful you are.

Agree 100%,the code in question is trivial and would have been faster to write then the original post.

[Deedle]

if you wonder "what will happen with this code", write the code and run it to find out.

Agree 100%,the code in question is trivial and would have been faster to write then the original post.

This can be useful as a learning exercise but IMO it's far far better to understand the language and other tools in use and write code that's correct and THEN verify it rather than determine by experiment what some code does until it seems to work. The latter can work in a pinch but if you're in a situation where there are a wide variety of possibly hostile inputs to your function, a variety of platforms it might find itself running on, and so on then you have a problem. All you've proved is it works for your test inputs on your specific platform.

The above discussion where no one considered rounding float values less than zero is a classic and extremely simplistic example.

[njineermike]

Just throwing this out there, but if forced to work in all INT, (and if the values allow it) you can multiply all values by 10 to get an implied decimal. Just bear in mind a signed 16 bit integer (WORD) limits you to 16383 and an unsigned to 32767, which imply 1638.3 and 3276.7. Multiply by 100 for 2 implied decimal places where values allow. Unfortunately, division and INT units are not friendly. You will always have the rounding error issue. But it's a great tool for addition, subtraction and multiplication with all INT.

I have to do that all the time when I'm stuck dealing with ancient technology. I then either shift the value on another function somehow or do the decimal point shift in a software format that allows it, like a graphics package. For example, a lot of analog input cards are INT only. Koyo cards are 12 bit, so I get 0-4096. An implied decimal gives me .1 unit resolution instead of INT unit resolution when measuring something that the span allows.