Error while calculating (4.25 - 5.35 equals 4.44089E-126)

[NorComm]

Greetings,

I have a rather unusual error. As you can see from the picture named "not_working.png", "x+y" should be zero, but instead it's "4.44089E-16". But, when I remove the "4.25 - 5.35" part of the calculation and do the math by hand (see "working.png") then everything works as it should. I simplified this .vi only to show you the mentioned problem. I am rather new to Labview, so I suppose the problem is in some minor detail, but I can't seem to find it.

Thank you in advance

[Ben]

Any time I see someone asking about epsilon and doing math with computers, I have to post a link to this thread.

The readers digest version of the reply amounts to this.

In the same way we can't express teh number 1/3 in decimal, the number 1/10 can not be expressed in binary.

See those links above for previous discusions on comparing numbers.

Ben