Broken equality function!?

Hey,

Im having problems with the equality function trying to compare vector angles, and I still can't understand what Im doing wrong. It almost seems to me as the equality (and actually <> functions too) are broken!? Clearly, index 1 in the marked paths in the picture should show "TRUE" as they are equal to index 1 in the list I am comparing the angles to? Im trying to single out path 0;1;24 in the picture, which is not aligned with either X or Y...

Seem really awkward that these simple math functions should be, but I simply can't see what Im doing wrong here..? Any help sorting this mess out would be greatly appriciated!!

Attachments:

equalityBroke.pdf, 253 KB

Replies to This Discussion

Permalink Reply by Jakob Lilliemarck on July 10, 2016 at 1:37am

Just solved this issue by using similarity instead of equality... given the circumstances its weird equality didn't work though, because the points that I draw the vectors from were taken from a rectangle in XY orientation, so most of them should be dead aligned with X or Y axis, but clearly they were not :S? Is there a way to make Grasshopper type all the decimals of a number or signal when something is not actually zero?

The panels really leave you in the dark here, especially when you can and should expect the numbers to be dead on...

Permalink Reply by David Rutten on July 10, 2016 at 6:42am

http://www.grasshopper3d.com/forum/topics/floating-point-error-why-...

Permalink Reply by David Rutten on July 10, 2016 at 6:45am

To be fair, it might be a bug but I cannot test without the data. What you can do is is use the format component to turn your numbers into exact (not rounded) textual representations. You'll need the formatting mask {0:R} and plug the numbers into the [0] input. (The 'R' flag stands for 'roundtrip', meaning the text can be converted back into the exact same number if need be, ie. the formatting is lossless).

Permalink Reply by Jakob Lilliemarck on July 10, 2016 at 9:02am

Hey David, thanks for your swift reply

Indeed, some of the numbers were rounded as can be seen in the picture, though I can't understand how, as all come from the same line intersection event, intersecting a rect in XY orientation with multiple curves. Anyway, good to learn how to reveal the real numbers in the panel!

Permalink Reply by David Rutten on July 10, 2016 at 11:03am

Floating point number mathematics is almost never exact. Whenever you add, multiply, subtract, or divide floating point numbers the result will be close to but not exactly what the answer in an ideal world would have been. Over time, these discrepancies add up and in some cases amplify.

RSS