A library providing basic functionality for Bidirectional Texture Functions
If you want to use the MATLAB code, you need to compile the bundled IEEE 754r Half Precision floating point library:
- change to the subfolder matlab/thirdparty/half_precision/
- from the MATLAB command window call halfprecision() without any arguments
- this should automatically call the mex compiler and if no errors are produced, you can test the library with
- halfprecision(halfprecision(single(0.1)), 'single')
- this should output 0.1 again
- you can now add the subfolder matlab/thirdparty/half_precision/ to your MATLAB search path: addpath('matlab/thirdparty/half_precision/') (from this repository's root folder)
The following changes have been applied to the original code from http://www.mathworks.de/matlabcentral/fileexchange/23173-ieee-754r-half-precision-floating-point-converter to make it compile on 64bit Linux systems:
-
in halfprecision.c add the following directives (replacing the existing #defines):
#include <stdint.h> #define INT16_TYPE int16_t #define UINT16_TYPE uint16_t #define INT32_TYPE int32_t #define UINT32_TYPE uint32_t -
in halfprecision.m replace "mex(cname);" with "eval(['mex -O CFLAGS="$CFLAGS -std=c99" ', cname]);"