You can always simulate multidimensional arrays using single dimensional arrays, but keep in mind that the Stamps have only 26 bytes of variable space available.
That depends on which Spin compiler you use or whether you use the Catalina C compiler. There are two 3rd party Spin compilers and one of them (Homespun) supports multidimensional arrays. The others (Parallax's Propeller Tool and BST) do not. They support only single dimensional arrays. The Propeller is fast enough and has enough RAM (32K) so simulating multidimensional arrays using single dimensional arrays is practical. As far as I know, Catalina supports multidimensional arrays. ImageCraft also has an excellent C compiler, but, although it's very inexpensive, it's not free like Catalina.
Bi-liner interpolation is the objective. More specifialy I would like to use a look-up table to calculate a pulsewidth, based off of a couple measured perimeters (say RPM and throdle postion).
Lookup tables on a Stamp are fairly easy. The table is kept in EEPROM where it's usually loaded with the program using DATA statements and accessed using READ statements. Typically the parameter tables are separate. You'd lookup the RPM and the throttle position in each of their tables and use their indices to compute an address in the pulsewidth table. Since you'd have the original RPM and throttle values as well as the parameter values at the found indices, you'd be able to do the interpolation on the pulsewidth.
Comments
You can always simulate multidimensional arrays using single dimensional arrays, but keep in mind that the Stamps have only 26 bytes of variable space available.
Why don't you describe what you're trying to do, and we'll see if it can be implemented on a BS2?
Dave