How to execute Python script inside a Matlab '' for loop ''?

I have two Matlab " for loops " like this,

%Matlab Code:

for M=1:NoRun
    for ii=1:NoP
           z1 = position(ii,1,M);        
           z2 = position(ii,2,M);

And inside the second " for loop " I have a ' Python script ' with variables "z1" and "z2" like this,

#Python Code:
        # Initialising the ANN
        model0 = Sequential()
        
        # Adding 1 hidden layer: the input layer and the first hidden layer
        model0.add(Dense(units = z1, activation = 'tanh', input_dim = 4, kernel_initializer=my_init))
        
        # Adding 2 hidden layer
        model0.add(Dense(units = z2, activation = 'tanh', kernel_initializer=my_init))
        
        # Adding output layer
        model0.add(Dense(units = 6, activation = 'softmax', kernel_initializer=my_init))
        
        # Set up Optimizer
        Optimizer = tf.train.AdamOptimizer(learning_rate=0.001, beta1=0.9, beta2=0.99)
        
        # Compiling the ANN
        model0.compile(optimizer = Optimizer, loss = 'categorical_crossentropy', metrics=['accuracy','categorical_crossentropy','mse'])
        
        # Fitting the ANN to the Train set, at the same time observing quality on Valid set
        history = model0.fit(X_train, y_train_dummy, validation_data=(X_test, y_test_dummy), batch_size = 100, epochs = 1500)
        
        # Generate prediction for all Train, Valid set and Experimental set
        y_train_pred_model0 = model0.predict(X_train)
        y_test_pred_model0 = model0.predict(X_test)
        y_exp_pred_model0 = model0.predict(E_data)
        
        # find final prediction by taking class with highest probability
        y_train_pred_model0 = np.array([[list(x).index(max(list(x))) + 1] for x in y_train_pred_model0])
        y_test_pred_model0 = np.array([[list(x).index(max(list(x))) + 1] for x in y_test_pred_model0])
        y_exp_pred_model0 = np.array([[list(x).index(max(list(x))) + 1] for x in y_exp_pred_model0])
        
        # check what metrics are in fact available in history
        history.history.keys()
        ac = (accuracy_score(y_test, y_test_pred_model0))*100

I need the value of accuracy_score 'ac' after each time executing the for loop, such that i can give this value to another variable inside this loop like this (written in Matlab)

%Matlab Cade:

Accurcy(M,ii)=ac;
           F1=(Accurcy);
           IAE(M,ii)=-F1;
           pBestFF(ii)=min(IAE(:,ii));
            if (pBestFF(ii)-IAE(M,ii))==0
                pBestX(ii,:)=position(ii,:,M);           %% Personal Best Particles
            end        
        end % Second for loop ends here.
    %% Global Best(gBest) Calculation
    gBestFF=min(pBestFF);
    for ii=1:NoP
        if (IAE(M,ii)-gBestFF)==0
        gBestX=position(ii,:,M);
        end
    end
        % Calculation of Particle Velocity and Position
        c1=2; c2=2;
        inertia(M)=((NoRun - M)*(0.9 - 0.4))/NoRun + 0.4;
           
        for ii=1:NoP
            for jj=1:dimension
                 velocity(ii,jj,M+1) =inertia(M)* velocity(ii,jj,M) + c1*rand(1)*(pBestX(ii,jj) - 
                 position(ii,jj,M)) + c2*rand(1)*(gBestX(jj) - position(ii,jj,M));%% Velocity update
                 velocity(ii,jj,M+1) = min(vmax(jj),max(vmin(jj),velocity(ii,jj,M+1)));
                 
                 position(ii,jj,M+1) = position(ii,jj,M) + velocity(ii,jj,M+1);
                 position(ii,jj,M+1)=min(smax(jj),max(smin(jj),position(ii,jj,M+1)));
                 
            end    
        end
        gbestFF1=[gbestFF1 gBestFF];
        Nrun1=[Nrun1 M];
    end % First for loop ends here.

I am doing this in Matlab. Suggest your Ideas for how to do this in Matlab.

Answer 1

[Bo Li]

Hopefully I understand the question correctly. I wonder you can create a Python function in a Python file mymodule.py to wrap your Python script for example myFunc like following:

#mymodule.py
#myFunc accepts z1 and z2 as inputs, returns ac as output
def myFunc(z1, z2):
    #Python logic
    return ac

Then you MATLAB loops could look like following:

for M=1:NoRun
    for ii=1:NoP
           z1 = position(ii,1,M);        
           z2 = position(ii,2,M);
           ac = py.mymodule.myFunc(z1, z2); #mymodule.py should be in Python PATH or current folder
           %use ac in more MATLAB logic

[N/A]

@Bo Li , Exactly this is what I mean. But I wonder how can I import Python script in matlab and pass the arguments (variables) to it and get the result from it, to pass the result to further Matlab Program.. Can you able to detail it.

[N/A]

@Bo Li, Iam getting an error : Undefined variable "py" or class "py.PythonScript.code"., while trying to execute the command "ac=py.PythonScript.code(140,49)'' in Matlab command prompt.

[Bo Li]

You will need to have CPython installed on your machine in order to invoke Python script from MATLAB. Depending on your MATLAB version, you might use either pyenv or pyversion to set up the Python interpreter:

https://www.mathworks.com/help/matlab/ref/pyenv.html

[Bo Li]

"how can I import Python script in matlab and pass the arguments (variables) to it and get the result from it, to pass the result to further Matlab Program"

%Say you create a Python file mymodule.py in folder "C:\work" and you have CPython 3.7 installed in "C:\python3" folder. You can launch MATLAB and do following in MATLAB:

>>cd c:\work
>>pyenv('Version', 'C:\python3\bin\python')
>>ac=py.mymodule.myFunc(1,2)  %py.mymodule imports the Python module in MATLAB