PDA

View Full Version : "Code efficient" floating point math


RFAndy
06-07-2010, 02:10 PM
Computing raw sensor values needs some sort of scaled down floating point math. Integer math is good but limits a designer/user to a limited range.

Case in point. The temperature sensors TMP35, MCP9700 and the humidity sensor HIH-5030 required a lot of sample averaging and math conversion needed
to get the analog result into usable engineering units.

I realize that floating point math would add a very large overhead to the SNAP python scripting. If having floating point is not possible then
having a "long" data type in SNAP will suffice.

TimWhite
07-30-2010, 01:02 PM
Right, I have the same issue with GPS math...need to be able to do distance to/bearing to calculations...