Basic Java Question

[LifeElevated 24.6k]

Given two double values double1 , double2, why isn’t it reliable to test their equality using:

double1 == double2;

???

Comment possible code below!!

If you are the first person to get this I will follow you!!

 

HINT:

Spoiler

If both doubles are assigned NaN  (not a number) it returns false! 

How could you COMPARE the two values against a real number (zero)

 

[FinnLord 532]

You can do

double1.equals(double2)

Never mind it's only for string

Updated May 17, 2018 by FinnLord
Wrong

[FinnLord 532]

1 hour ago, LifeElevated said:

How could you COMPARE the two values against a real number (zero)

Can you specify it?... I dont understand

is it this?

Double.compare(double1,double2) == 0);

Updated May 17, 2018 by FinnLord

[Crypto 7.7k]

Updated May 17, 2018 by Crypto

[Crypto 7.7k]

18 hours ago, FinnLord said:

Can you specify it?... I dont understand

is it this?

Double.compare(double1,double2) == 0);

Yes this is the right answer. Now tell us why (==) isn't efficient to compare double variables?

[FinnLord 532]

5 hours ago, Crypto said:

Yes this is the right answer. Now tell us why (==) isn't efficient to compare double variables?

idk lol.....

[LifeElevated 24.6k]

6 hours ago, Crypto said:

Yes this is the right answer. Now tell us why (==) isn't efficient to compare double variables?

Honestly this is worded to weird ?

[Guest]

On 5/18/2018 at 12:21 AM, LifeElevated said:

Honestly this is worded to weird ?

Its not. I joined this club so i could respond

 

If you have a double or a float in any language, you can't expect them to be completely precise. For example, if you have two doubles 

double num1 = 3.0;
double num2 = 3.0;

It looks like they should be equal, right? But they most likely aren't. why? because  floats and doubles are very precise. Here's what num1 and num2 really are:

num1: 3.00000000000000000000000000000000382918
num2: 3.00000000000000000000000000000000261527

So that's why you can't compare doubles and floats with the == operator in any language really. Here's some code to correctly compare floats in c++

bool cmpf(float a, float b){
	return (fabs(a - b) < 0.005f);
}

I guess in Java it would be something ridiculous like

static boolean cmpf(double a, double b){
	return (Math.abs(a - b) < 0.005);
}

I f***ing hate java