With your first call of load:
load -mat filename
matlab will try to load a file with the explicit filename filename (or filename.mat). regardless of what the variable filename is. load on that line is run as the command load, the additional text on that line is taken as litteral strings. This is very convenient when running stuff from the command-line (and for scripts with fixed inputs), but confusing when converting to general-purpose functionality.
If you remove that line your code should work. You could consider changing:
load('fileName')
to:
load('-mat','fileName')
and
'File did not find';
to
disp(['Could not find file: ',filename]);
Also when asking about help with error-hunting, you get better replies if you include the full error-message.
(and cut the clear all, it's uggly)
HTH
