ppval problems with multidimensional spline

I am using csapi to fit a piecewise polynomial to a set of 3d data (2 inputs, 1 response). The code used to create the fit is the following:

---------------------------------

 cL_pp = csapi( {aoa, mach}, cL );

---------------------------------

where:

aoa is a 100x1 vector

mach is a 80x1 vector

cL is a 100x80 matrix

Later on in my code, when I try to use cL_pp to predict a new cL value, I get an error. The code and error are below:

---------------------------------

 current_cL = ppval( cL_pp, [current_aoa current_mach] );

??? Index exceeds matrix dimensions.

Error in ==> ppval at 67 if lx, [~,index] = histc(xs,[-inf,b(2:l),inf]);

---------------------------------

I have tried several variations to my call to ppval using curly brackets and square brackets; changing the prediction site to be 2x1 instead of 1x2, etc. Trial and error isn't the most efficient way to figure out proper syntax though...

I dug into ppval.m to try to find a solution, and the error looks suspiciously like a coding error in ppval itself, but the indexing used may just be over my head. I am loathe to suggest that I could code better than Matlab :)

Anyone that can point out the error of my ways would be very much appreciated!

p.s. I'm using R2011a