When does round(0.0)=NaN?

General
1

Ran into this earlier and can’t figure out how round(0.0) returns NaN.

I have a structure, pnt{x as double, y as double}
2 lines
pnt.x=round(pnt.x)
pnt.y=round(pnt.y)
somehow I end up with {NaN,0}

I changed the code to assign to an integer as the rounding step, but I’d like to understand what might cause this… any ideas?

2

Perhaps try testing to see that pnt.x and pnt.y actually have values, or set them to a default value of 0.0?

3

They show as 0 in the debugger. I’m wondering if there’s something strange in the calculations that arrive at zero… other values have no issue. and zero only has trouble in the x calculation…

4

32-Bit or 64-Bit?