2

I want to call some python code from MATLAB, in order to do this I need to convert a matrix object to a NumPy ndarray, through the MATLAB function py.numpy.array. However, passing simply the matrix object to the function does not work. At the moment I solved the problem converting the matrix to a cell of cells object, containing the rows of the matrix. For example

function ndarray = convert(mat)
    % This conversion fails
    ndarray = py.numpy.array(mat)

    % This conversion works
    cstr = cell(1, size(mat, 1));
    for row = 1:size(mat, 1)
        cstr(row) = {mat(row, :)};
    end
    ndarray = py.numpy.array(cstr);

I was wondering if it exists some more efficient solution.

Improve this question
5
  • Keep in mind you can only send vector from Matlab to Python. 1XN In file.py convert it to a numpy ndarray. Commented Jan 17, 2017 at 19:56
  • Which versions of MATLAB and Python are you using? What are the dimensions of the array? Please also provide a minimal reproducible example in your question. Commented Jan 17, 2017 at 20:03
  • @TonyTannous actually I think that using cell of cells is correctly interpreted in numpy as a MxN ndarray. Commented Jan 17, 2017 at 20:05
  • @Dev-iL I am using MATLAB 2016a and python3.5. The array can become very large, thus if I need to convert the matrix object to an intermediate structure I want to make a time efficient conversion. Commented Jan 17, 2017 at 20:08
  • "Large" is meaningless. I meant - how many dimensions does it have? What are typical and maximal sizes (perhaps there's some memory limitation to the MATLAB-Python interface)? In any case I think you could reshape the array to a vector in MATLAB, then reshape it back in numpy. Commented Jan 17, 2017 at 20:09

2 Answers 2

Reset to default
3

Assuming your array contains double values, the error tells us exactly what we should do:

A = magic(3);
%% Attempt 1:
try 
npA = py.numpy.array(A);
% Result:
%   Error using py.numpy.array
%   Conversion of MATLAB 'double' to Python is only supported for 1-N vectors.
catch
end
%% Attempt 2:
npA = py.numpy.array(A(:).');
% Result: OK!

Then:

>> whos npA
  Name      Size            Bytes  Class               Attributes

  npA       1x1                 8  py.numpy.ndarray

Afterwards you can use numpy.reshape to get the original shape back, either directly in MATLAB or in Python.

Improve this answer
Sign up to request clarification or add additional context in comments.

3 Comments

Actually, I think this is the most efficient solution, but it needs to reshape the ndarray in python, and this is not always directly possible. If the reshape is not possible I think that my solution is better.
I have to correct my previous statement, I did not think that the NumPy reshape can be done directly in MATLAB through npA = npA.reshape(size(A,1), size(A,2)).
@AreTor Thanks. I've updated the answer to make this possibility clear.
0

Actually, using python 2.7 and Matlab R2018b, it worked with simply doing:

pyvar = py.numpy.array(var);

Matlab tells me that if I want to convert the numpy array to Matlab variable, I can just use double(pyvar)

By the way, it didn't worked with python 3.7, neither using an older version of Matlab . I don't know what this means, but I thought this might be helpful

Improve this answer

Comments

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.