problems when using scipy.optimize.differential_evolution in Matlab environment

Question

Reinhardt RADING 2023-6-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1977284-problems-when-using-scipy-optimize-differential_evolution-in-matlab-environment

编辑： Walter Roberson 2023-6-4

在 MATLAB Online 中打开

When i call a matlab function to be optimized with the inputs arguments i get 'not enough input' error.

For example; i have a matlab function called Residual.m that requires 5 input and returns one value.

Residual(state, N, freq, SNR, p)

My code is like this:

py_optimize = py.importlib.import_module('scipy.optimize'); 
N=2 
bound = [-50, 500; -50, 500; -150, 150]; 
x_bound = repmat(bound, N, 1); 
freq=10; SNR=30; 
p=1; 
data=py_optimize.differential_evolution(Residual,x_bound,disp=True, maxiter=4500, popsize=65, init='random', workers=worker, recombination=0.8, mutation=nn, strategy='best1bin', tol=0.001, updating='deferred'); 

When i run this, i get an error that there is not enough inputs.

I tried using matlabFunction with my fuction but i keep getting below error

Error using indexing

Handle to MATLAB function '@()' is not supported. Use a handle to a Python function.

What I'm i not doing right?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Al Danial 2023-6-4

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1977284-problems-when-using-scipy-optimize-differential_evolution-in-matlab-environment#answer_1249869

scipy.optimize.differential_evolution() expects a Python function as its first argument. You'll need to translate Residual.m to Python, then pass this translated function as the first argument. I suppose in theory you could provide a Python wrapper to Residual.m then use the matlab.engine from Python to call Residual.m but I've not been successful with that combination.