Trying to make an app in app designer to predict using trained NN model from classification learner but fails to classify the data

I managed to import data in csv and show them in the application table, I manage to load the model with the second button but when I try to predict the data I get an error saying that it does not recognize the function I use to predict the data normally with the model.

I have tried to predict using the way they advise in other cases:

Ypred = predict(model, data);

but I got the error:

Error using app1/predictButtonPushed

Undefined function 'predict' for input arguments of type 'table'.

I suspect the latter is due to the fact that the neural network is stored as a structure where the function used to make the prediction is a function handle.

here is the code of the app so far:

classdef app1 < matlab.apps.AppBase

    % Properties that correspond to app components
    properties (Access = public)
        UIFigure             matlab.ui.Figure
        predictButton        matlab.ui.control.Button
        ImportmodelButton_2  matlab.ui.control.Button
        ImportdataButton     matlab.ui.control.Button
        UITable2             matlab.ui.control.Table
        UITable              matlab.ui.control.Table
    end

    
    properties (Access = private)
        data_imp % Description
        pred_NN2lys
    end
    

    % Callbacks that handle component events
    methods (Access = private)

        % Button pushed function: ImportdataButton
        function ImportdataButtonPushed(app, event)
            % Load the trained classifier
            [filename, filepath] = uigetfile({'*.csv'; '*.xlsx'}, 'Select Data File');
            data = readtable(fullfile(filepath, filename));
            app.data_imp = data;
            app.UITable.Data = data;          
        end

        % Button pushed function: ImportmodelButton_2
        function ImportmodelButton_2Pushed(app, event)
           [filename, filepath] = uigetfile('*.mat', 'Select Data File');
           NN2lys = load(fullfile(filepath, filename));
           app.pred_NN2lys = NN2lys                    
        end

        % Clicked callback: UITable
        function UITableClicked(app, event)
            displayRow = event.InteractionInformation.DisplayRow;
            displayColumn = event.InteractionInformation.DisplayColumn;
            
        end

        % Button pushed function: predictButton
        function predictButtonPushed(app, event)
            input = app.data_imp;
            YPred = app.pred_NN2lys.predictFnc(input);           
        end
    end

    % Component initialization
    methods (Access = private)

        % Create UIFigure and components
        function createComponents(app)

            % Create UIFigure and hide until all components are created
            app.UIFigure = uifigure('Visible', 'off');
            app.UIFigure.Position = [100 100 640 480];
            app.UIFigure.Name = 'MATLAB App';
            app.UIFigure.Scrollable = 'on';

            % Create UITable
            app.UITable = uitable(app.UIFigure);
            app.UITable.BackgroundColor = [1 1 1;0.9412 0.9412 0.9412];
            app.UITable.ColumnName = {'Ag'; 'Bi'; 'Cu'; 'Fe'; 'Ni'; 'Pb'; 'Pd'; 'Pt'; 'Sb'; 'Te'};
            app.UITable.RowName = {};
            app.UITable.ColumnEditable = true;
            app.UITable.ClickedFcn = createCallbackFcn(app, @UITableClicked, true);
            app.UITable.Position = [16 31 408 419];

            % Create UITable2
            app.UITable2 = uitable(app.UIFigure);
            app.UITable2.ColumnName = {'Deposit'};
            app.UITable2.RowName = {};
            app.UITable2.ColumnEditable = true;
            app.UITable2.Position = [554 31 75 419];

            % Create ImportdataButton
            app.ImportdataButton = uibutton(app.UIFigure, 'push');
            app.ImportdataButton.ButtonPushedFcn = createCallbackFcn(app, @ImportdataButtonPushed, true);
            app.ImportdataButton.Position = [436 272 100 23];
            app.ImportdataButton.Text = 'Import data';

            % Create ImportmodelButton_2
            app.ImportmodelButton_2 = uibutton(app.UIFigure, 'push');
            app.ImportmodelButton_2.ButtonPushedFcn = createCallbackFcn(app, @ImportmodelButton_2Pushed, true);
            app.ImportmodelButton_2.Position = [436 230 100 23];
            app.ImportmodelButton_2.Text = 'Import model';

            % Create predictButton
            app.predictButton = uibutton(app.UIFigure, 'push');
            app.predictButton.ButtonPushedFcn = createCallbackFcn(app, @predictButtonPushed, true);
            app.predictButton.Position = [436 184 100 23];
            app.predictButton.Text = 'predict';

            % Show the figure after all components are created
            app.UIFigure.Visible = 'on';
        end
    end

    % App creation and deletion
    methods (Access = public)

        % Construct app
        function app = app1

            runningApp = getRunningApp(app);

            % Check for running singleton app
            if isempty(runningApp)

                % Create UIFigure and components
                createComponents(app)

                % Register the app with App Designer
                registerApp(app, app.UIFigure)
            else

                % Focus the running singleton app
                figure(runningApp.UIFigure)

                app = runningApp;
            end

            if nargout == 0
                clear app
            end
        end

        % Code that executes before app deletion
        function delete(app)

            % Delete UIFigure when app is deleted
            delete(app.UIFigure)
        end
    end
end

[Christopher McCausland]

Just a comment,

If you are using a line like this;

predict = predict(model, data);

MATLAB, or any other programing langauge will error as 'predict' is now both a function and a variable. This type of overshadowing is handled by matlab in that the function cannot run until the variable is cleared from memory. You can read more about function precedence order here.

Christopher

[Angel Verbel]

Hi Christopher,

My mistake, I changed the name of the variable when I copied it here because it was originally in Spanish, so they are different in the original scrip.

Thank you for bringing this point to my attention.

Angel

Answer 1

[Githin George]

Hello,

My understanding is that you are having a model exported from the Classification Learner app named ‘NN2lys’ and you are facing an issue when using the “predictFcn” method of this model in “predictButtonPushed” callback.

The “predictFcn” requires the data with the same format and data type as the training data used in the Classification Learner app. Use the following command on MATLAB command window to get the right instructions on how to pass data to the “predictFcn” method.

NN2lys.HowToPredict

As an alternative, the struct ‘NN2lys’ also contains the actual trained model which can be used directly with the “predict” function.

PS: There is a typo in Line 2 of the “predictButtonPushed” callback. Please change it to the following:

YPred = app.pred_NN2lys.predictFcn(input);

I hope this helps.

Answer 2

[Federico Mazzoli]

Hi Angel,

I have the same problem.

You solved it somehow?

Federico