C fixed point calculation

[robert2032survey]

G'day all

I am building a calculation application. The design of the application is made in matlab. I converted the script into Perl. Because of performance issues Perl was not quick enough. In the Perl application I included Inline::C. Now the application is fast enough, but the calculations are wrong compared to Matlab and Perl. I have found out that this is caused by the difference in data types.

My question:
How do I correct this? So the calculation is equal between Matlab/Perl and C.

My application performs several Matrix operations. Are there any C-matrix library's out there?

Best Regards Robert

[horace1]

you need to find out how Perl and the C compiler you are using represent the data types (number of bytes, etc) and see if you can find a common data format to pass information between the two. Otherwise you will have to do some translation as data is passed. This is also a problem when passing data between C/C++ programs implemented using different compilers even on the same machine, e.g. one compiler may represent an int with 16 bits another may use 32 bits, different machines may be little endian or big endian, etc.
For a matrix library have a look at
http://www.gnu.org/software/gsl/