mxCreateNumericArray error: cannot convert 'int*' to 'const size_t* {aka const long long unsigned int*}' for argument '2' to 'mxArray*

I am trying to replicate the paper of Druedahl and Jorgensen (2017), with title "A general endogenous grid method for multi-dimensional models with non-convexities and constraints". However, I have an issue with the compiler in matlab, to read C++ files. More specifically, when I do in matlab:

mex cfuncs\mexUpperEnvelopeToCommon.cpp CXXFLAGS="$CXXFLAGS -fopenmp -std=c++11 -O3 -march=native -ffast-math -fabi-version=0 -mavx2" C:\TDM-GCC-64\lib\gcc\x86_64-w64-mingw32\5.1.0\libgomp.a;

I get the error:

error: cannot convert 'int*' to 'const

size_t* {aka const long long unsigned int*}' for argument '2' to 'mxArray* mxCreateNumericArray_730(size_t, const

size_t*, mxClassID, mxComplexity)'

plhs[iout] = mxCreateNumericArray(par->ndim,dims,mxLOGICAL_CLASS,mxREAL);

Also mex -DMX_COMPAT_32 cfuncs\mexUpperEnvelopeToCommon.cpp gives the same error.

I already tried to replace all int with mwSize. However, there are also two int* in the file, and I do not know how to replace those.

Does anyone know how to solve the issue? I attach the zip file with the C++ functions.

Answer 1

[James Tursa]

Simply replace this

int* dims = new int[par->ndim];

with this

mwSize* dims = new mwSize[par->ndim];

[Dick Oosthuizen]

Then I get the following error:

cfuncs\mexUpperEnvelopeToCommon.cpp:223:77: error: cannot convert 'mwSize* {aka long

long unsigned int*}' to 'int*' for argument '3' to 'void mesh::create(mesh::settings_struct*, int, int*, double**,

double*)'

mesh::create(w_interp_func, par->ndim, grid_lengths, grid_vectors, w);

[James Tursa]

So, how did changing the dims definition affect grid_lengths? Did you incorrectly change grid_lengths to mwSize* when you should have left it as int*?

[Dick Oosthuizen]

I fixed that, then i get the error:

Error using mex

Cannot export mexFunction: symbol not defined

collect2.exe: error: ld returned 1 exit status

[James Tursa]

Please show the complete mex command you are using and the resulting error. I clearly see mexFunction in your mexUpperEnvelopeToCommon.cpp file, so I don't understand the error yet.

[Dick Oosthuizen]

I run:

mex cfuncs\mexUpperEnvelopeToCommon.cpp XXFLAGS="$CXXFLAGS -fopenmp -std=c++11 -O3 -march=native -ffast-math -fabi-version=0 -mavx2" C:\TDM-GCC-64\lib\gcc\x86_64-w64-mingw32\5.1.0\libgomp.a;

I get:

Building with 'MinGW64 Compiler (C++)'.

Error using mex

Cannot export mexFunction: symbol not defined

collect2.exe: error: ld returned 1 exit status

[Dick Oosthuizen]

I found my error! It had to do with the MexFunction. Thank you so much!

[晓航 邵]

Do you finally use this kinof commamd "mex -compatibleArrayDims readTrackletsMex.cpp".

But why it does work in my situation? Could you tell me how to solve it?

errors still occurred like:

使用 'g++' 编译。

错误使用 mex

/home/shawn/Documents/MATLAB/StereoKittiSLAM/utilsRawData/readTrackletsMex.cpp: In function ‘void mexFunction(int, mxArray**, int, const

mxArray**)’:

/home/shawn/Documents/MATLAB/StereoKittiSLAM/utilsRawData/readTrackletsMex.cpp:76:82: warning: narrowing conversion of

‘tracklets->Tracklets::getTracklet(tracklet_idx)->Tracklets::tTracklet::poses.std::vector<Tracklets::tPose>::size()’ from

‘std::vector<Tracklets::tPose>::size_type {aka long unsigned int}’ to ‘int’ inside { } [-Wnarrowing]

const int poses_dim[] = {15, tracklets->getTracklet(tracklet_idx)->poses.size()};

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~

/home/shawn/Documents/MATLAB/StereoKittiSLAM/utilsRawData/readTrackletsMex.cpp:108:3: error: ‘cout’ was not declared in this scope

cout << "Found " << tracklets->numberOfTracklets() << " tracklets." << endl;

^~~~