Wrong Values of Doubles from .mat file in C

I have a C program reading a .mat file. The file contains 2 1d arrays of doubles. The program gets a pointer to the beginning of an array of doubles by using mxGetDoubles(). It can successfully read all the values, however there appears to be a scaling problem when printing them with printf("%lf\n", ...) , e.g.

13.69798 displays as as 24.553575

48.94982 displays as as 57.016571

Matlab is R2022a

GCC is x86_64 V11.3.0 (MINGW running in Windows)

Could there be a difference in encoding? I thought they would both use IEEE 754 double encoding.

[Walter Roberson]

?? mxGetDoubles() is not (in itself) suitable for reading doubles from a .mat file.

Have you tested with https://www.mathworks.com/help/matlab/matlab_external/reading-a-mat-file-in-cc.html ?

[Tony Whittam]

Perhaps I should have given more detail. Here is the C code ...

    const char *file1 = "matlab/test/testData/testfile.mat";
    MATFile *pmat;
    mxArray* pa;
    mxArray* pa1;
    mxDouble* extData;
        
    pmat = matOpen(file1, "r");

    pa1 = matGetNextVariableInfo(pmat, &name);
    pa = matGetVariable(pmat, name);
    extData = mxGetDoubles(pa);

    for (i=0; i<32; i++)
    {
         printf("%lf\n", extData[i]);
    }

Answer 1

[Tony Whittam]

I just solved our problem. Operator error as usual. We were looking at different data files in Matlab and the C code. Once I opened the same file in Matlab, the numbers were identical to those printed in C.