Question about coordinate problem at zero

Pixero

Hi!

I have a question that is not directly about SketchUp or Ruby but still a problem...

I need coordinate values that are positive so I use abs(x) to get it from negative coordinates and that works fine.
The problem I have is that when the coordinates are precisely at 0, (x==0 || y==0 || z==0) the rest of the math doesn't work.
My experiments with it havent worked so far.
For example I've tried with setting it to a very small value if the coordinate is exactly 0.
I'm unsure what to Google for to find any info.

Does anyone know how these things are usually solved?

(Sorry for beeing a bit cryptic and a bit off topic.)

dukejazz

aspect ratio of some objects won't center on [0,0,0] with snap
usealy two coordinate ok one not
try turning off snap

it still dosn't work
you might need more precision.
make sure you object is a group
it gives 2x precision
but don't expode complety if your group object smaller .001, it will disapear from view
hit undo if it doses.

here a plugin that will shape a object to a better precision
scaleGroupFloat_dj Ver 1.6.5
Copyright Feb 15, 2013
DukeJazz: by James Cochran
http://sketchucation.com/forums/viewtopic.php?f=323&t=45254

Info: Best high precision 3d scale and move tool yet.

add zoro to numbers of lenght
1.0000001

jolran

Hej.

Isent abs kind of slow ?
I used it before for something for the same reason as you, but got very slow code .

A hack would be to round of to around 0,001 or something. Might avoid floating point precision errors as well.

Edit: Sorry did not read you post thourougly.

@unknownuser said:

For example I've tried with setting it to a very small value if the coordinate is exactly 0.

thomthom

Difficult to say what to do without knowing what the purpose of the code is. Sounds like 0 is an edge case for you - if you cannot make the calculation work with 0, then maybe you need to make a special processing case for when you're values are 0.

Pixero

Thanks for your replies.
This isnt directly connected to SU and is from a C++ project of mine.
However I solved it.
Turned out I needed a custom fmod function since the one built into C really is a remainder function and not a modulo function.