Float Question

CannibalRobotics

Why does:

TimeZone := f.Ftrunc(f.div(Lon,15.0))

give me the result: -6.085645e+38 when sent to debug with

debug.str(fs.FloatToString(TimeZone))

'Lon' has been set up as a floating point value via FFloat() and is correctly displayed as -97 with

debug.str(fs.FloatToString(Lon))

The -6 part is correct but neither FRound or FTrunc seem to work at 'intgerizing'.
I'm missing something...

John Abshier

TimeZone is an integer.
Replace debug.str(fs.FloatToString(Lon)) with debug.dec(TimeZone)

John Abshier

CannibalRobotics

Well, hum.
It's interesting that fRound and fTrunc return integers but all of the other functions return float's,,, I guess - it does leave me wanting a better documentation reference to the Float32 routines.
Any suggestions? I looked at Cam's write up and it's great on time and space but still does not say much for specific call functionality.
Thanks in advance.

lonesock

If you want it returned as a float, I added a "FloatTrunc" function to F32 (see my sig). The documentation is different, though I wouldn't necessarily say "better" [8^).

Jonathan

Mike Green

fRound and fTrunc are specific conversion routines for floating point to integer just like fFloat goes from integer to floating point. There happen to be two separate calls because they are two commonly used choices of what to do with the fractional part of the floating point number, round to the nearest integer or truncate the fraction. They probably didn't get much coverage in the documentation because they're so common in floating point libraries and instruction sets.