Best way for check if a variable is a whole number

Question

I need to check if a variable is a whole number and i have a lot of easy solutions.

The first one is

if (floor(foobar) == foobar)

And the other one is

if(d % 1 == 0)

The another one is convert this number to string and split by dot(.) and check second number is "00" of not

d.toString().split('.')[1/]=="00"

But i think it is not a best practice.

A lot of these type of way, i want a standerd solution which is acceptable by all.

There is actually some discussion in the comments on the answers you linked about this.. — Alex Booker, Aug 11 '15 at 09:52
possible duplicate of [How to determine if a decimal/double is an integer?](http://stackoverflow.com/questions/2751593/how-to-determine-if-a-decimal-double-is-an-integer) — Nahum, Aug 11 '15 at 09:55
You cannot choose if you don't know what you need. Write down your exact requirements and validate whether either fulfills all requirements. If both do, roll a die. — CodeCaster, Aug 11 '15 at 09:56
`(int)d == d` will return false if d is outside the integer-range — Dennis_E, Aug 11 '15 at 09:56
Considering that the first answer is about C++, perhaps you have bigger problems than choosing the best one... — xanatos, Aug 11 '15 at 09:57
@Binkan Salaryman. Then it will fail if d is outside the long-range. — Dennis_E, Aug 11 '15 at 10:00
Don't try to learn billions of "best" ways to write code. Write clear, understandable code and set performance goals. Then measure the performance. If the performance is unacceptable, isolate *where* the code is performing badly and look to optimize there. — Damien_The_Unbeliever, Aug 11 '15 at 10:00
if type of variable d is Double, then what about "(int)d == d" — mayank, Aug 11 '15 at 10:01
@Dennis_E stop being pedantic, you got the idea how it should work — Binkan Salaryman, Aug 11 '15 at 10:01
`Best means "Fast in execution" and "Less Load on Server"`. This is probably premature optimization. Do you think this code will be called so many times that the performance will suffer. If yes, have you measured it? — publicgk, Aug 11 '15 at 10:01
@BinkanSalaryman But it doesn't work. That's the whole point. — Dennis_E, Aug 11 '15 at 10:02

score 9 · Answer 1 · answered Aug 11 '15 at 10:03

9

I'd probably go with

(d % 1) < epsilon

from the comment thread in the linked question. It's slower than both your examples so you can probably only do it a few million times per second as opposed to a few million times per second. On the plus side it handles floating point drift and min and max values quite well

answered Aug 11 '15 at 10:03

James

9,774
5
34
58

2

+1 for "a few million as opposed to a few million". Premature optimization is indeed the root of all evil. – CodeCaster Aug 11 '15 at 10:04

score 0 · Answer 2 · answered Aug 11 '15 at 09:53

0

IMO, (int)d == d is the best in regards to readability.

I don't know about execution times, though.

answered Aug 11 '15 at 09:53

Nathan

1,520
1
12
21

2

The `int.MaxValue` is quite small if compared to `double.MaxValue` – xanatos Aug 11 '15 at 09:56

score 0 · Answer 3 · answered Apr 19 '20 at 05:21

0

In C# you can use built in function int.TryParse()

int x;
bool result= int.TryParse(integerNumber.ToString(), out x);

if result is true then integer otherwise not.

answered Apr 19 '20 at 05:21

kode_anil

66
1
6

Not quite. Not all whole number `doubles` are valid integers. `int.TryParse((int.MaxValue + 1.0).ToString(), out _)` return `false`! – Pharaz Fadaei Sep 01 '21 at 21:28

Best way for check if a variable is a whole number

3 Answers3