Please note that the recommended version of Scilab is 2024.0.0. This page might be outdated.
See the recommended documentation of this function
mexlib
Mexlib provides a C MEX API for Scilab. This module is based on the MEX specification.
How to
To use the MEX API you first need to compile a C file. Here is a small example (call it example.c):
#include "mex.h" void mexFunction(int nlhs, mxArray *plhs[], int nrhs, mxArray *prhs[]) { double *output; double data[] = {1.0, 2.1, 3.0}; plhs[0] = mxCreateDoubleMatrix(1, 3, mxREAL); output = mxGetPr(plhs[0]); memcpy(output, data, 3*sizeof(double)); }
This function just returns a double array with 3 values. The important part is the MEX API is used; mex.h is included and mexFunction defined with the right calling convention and arguments.
To compile this file, you should use Scilab ilib_mex_build function.
--> ilib_mex_build('libmexexample',['mexgetdouble','example', 'cmex'], 'example.c',[],'Makelib','','','');
This creates a loader file. To load it into Scilab, you should run:
--> exec('loader.sce');
List of different mex function categories
- Data Types
Data types and constants.
- Create or Delete Array
Create array of a particular type, allocate and free memory.
- Validate Data
Evaluate array data type.
- Access Data
Reading or writing data to an array.
- Convert Data Types
Convert strings and structure array to object array.
Examples
#include "mex.h" void mexFunction(int nlhs, mxArray *plhs[], int nrhs, mxArray *prhs[]) { double *output; double data[] = {1.0, 2.1, 3.0}; plhs[0] = mxCreateDoubleMatrix(1, 3, mxREAL); output = mxGetPr(plhs[0]); memcpy(output, data, 3*sizeof(double)); }
After that, just call your function, created with name mexgetdouble:
--> ilib_mex_build("libmex",["mexgetdouble","mexgetdouble.c","cmex"], "mexgetdouble.c",[]); --> exec("loader.sce"); --> r = mexgetdouble() r = 1 2.1 3
#include "mex.h" void mexFunction(int nlhs, mxArray *plhs[], int nrhs, mxArray *prhs[]) { mexPrintf("\nThis is an example Mex function.\n"); int n = 0; mxArray* pOut = NULL; n = mxGetNumberOfDimensions(prhs[0]); pOut = mxCreateDoubleScalar((double)n); plhs[0] = pOut; }
Assuming that the file name is mexexample.c, just call your function, created with the name mexexamp:
--> ilib_mex_build("libmex",["mexexamp","mexexample.c","cmex"], "mexexample.c",[]); --> exec("loader.sce"); --> a = ones(1, 2, 3, 4); --> mexexamp(a) This is an example Mex function. ans = 4.
See also
- ilib_mex_build — utility for mex library management
- MEX-files guide from Matlab
- MEX-files in the GNU Octave manual
History
| Version | Description |
| 6.1 | Improved MEX API coverage. |
| 5.0 | MEX API initial support. |
| Report an issue | ||
| << Validate Data | MEX Library API | Advanced functions >> |