Display output from neural network on APP Designer

Question

Aditi Mahajan 2023-4-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1953934-display-output-from-neural-network-on-app-designer

评论： Aditi Mahajan 2023-5-5

I have build a neural network model and saved it to load in APP Designer as shown in figure.

The code is showing no error. But also, it doesn't display the output as required. I have singularly checked each code line's execution in live script and all is working fine. The saved model (MSB.mat) is also being loaded and working fine.

What could be the possibility of no output?

Kindly help!

  % Callbacks that handle component events
    methods (Access = private)
        % Button pushed function: CheckforsuitablematrixButton
        function CheckforsuitablematrixButtonPushed(app, event)
        
%        Input variables from the user (Process,Country,FTr,FWf,FAR,TS,TM,FS,FM)
            Process=[app.ManufacturingProcessDropDown.Value]
            Country=[app.GeographicalLocationDropDown.Value]
            FWf=[app.FiberWeightFractionEditField.Value]
            FAR=[app.FiberAspectRatioEditField.Value]
            TS=[app.TensileStrengthMPaEditField.Value]
            TM=[app.TensileModulusGPaEditField.Value]
            FS=[app.FlexuralStrengthMPaEditField.Value]
            FM=[app.FlexuralModulusGPaEditField.Value]
            if strcmp(app.FiberTreatmentDropDown.Value,'Untreated')
                FTr=[0];
            else
                FTr=[1];
            end
%        Aggregating input variables into a table named IP
            IP=table(Process,Country,FTr,FWf,FAR,TS,TM,FS,FM)
%        Loading the trained neural network
            load MSB.mat
%        Predicting the output from the input
            Pred_M=predict(MSB,IP)
%        Displaying the output 
            app.UsethegivenMatrixEditField.Value=Pred_M
        end
    end

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

VBBV 2023-4-27

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1953934-display-output-from-neural-network-on-app-designer#answer_1224504

编辑：VBBV 2023-4-27

在 MATLAB Online 中打开

%Loading the trained neural network
          mdl = load('MSB.mat')
%        Predicting the output from the input
  Pred_M=predict(mdl,IP.Process)

Check the syntax for using the predict function

12 个评论
显示 10更早的评论隐藏 10更早的评论

VBBV 2023-4-27

在 MATLAB Online 中打开

Ok, Since Pred_M seems a categorical variable , use string(Pred_M) to display the output, and also see the following changes to display table data. Hope this helps

 % Callbacks that handle component events
    methods (Access = private)
        % Button pushed function: CheckforsuitablematrixButton
        function CheckforsuitablematrixButtonPushed(app, event)
        
%        Input variables from the user (Process,Country,FTr,FWf,FAR,TS,TM,FS,FM)
            Process=[app.ManufacturingProcessDropDown.Value]
            Country=[app.GeographicalLocationDropDown.Value]
            FWf=[app.FiberWeightFractionEditField.Value]
            FAR=[app.FiberAspectRatioEditField.Value]
            TS=[app.TensileStrengthMPaEditField.Value]
            TM=[app.TensileModulusGPaEditField.Value]
            FS=[app.FlexuralStrengthMPaEditField.Value]
            FM=[app.FlexuralModulusGPaEditField.Value]
            if strcmp(app.FiberTreatmentDropDown.Value,'Untreated')
                FTr=[0];
            else
                FTr=[1];
            end
         
            % use a UIFigure and UItable to display data in Apps
           fig = uifigure            
    %        Aggregating input variables into a table named IP
            IP=table(Process,Country,FTr,FWf,FAR,TS,TM,FS,FM,'VariableNames',{'Process','Country','FTr','FWf','FAR','TS','TM','FS','FM'});
            UIT = uitable(fig,'Data',IP)
%        Loading the trained neural network
            load MSB.mat
%        Predicting the output from the input
            Pred_M=predict(MSB,IP)
%        Displaying the output------------------>>> use a string to display
            app.UsethegivenMatrixEditField.Value  =  string(Pred_M)
        end
    end

VBBV 2023-4-28

在 MATLAB Online 中打开

You also need to assign the values to variables in the table as shown below and display using UIFigure in the App

% capture event by user input
Process = event.ManufacturingProcessDropDown.Value;
% assign the value to variable 
app.ManufacturingProcessDropDown.Value = Process;
Country = event.GeographicalLocationDropDown.Value
app.GeographicalLocationDropDown.Value = Country;
FWf = event.FiberWeightFractionEditField.Value
app.FiberWeightFractionEditField.Value = Fwf
FAR = event.FiberAspectRatioEditField.Value
app.FiberAspectRatioEditField.Value = FAR
TS = event.TensileStrengthMPaEditField.Value
app.TensileStrengthMPaEditField.Value = TS
TM = event.TensileModulusGPaEditField.Value
app.TensileModulusGPaEditField.Value = TM
FS = event.FlexuralStrengthMPaEditField.Value
app.FlexuralStrengthMPaEditField.Value = FS
FM = event.FlexuralModulusGPaEditField.Value
app.FlexuralModulusGPaEditField.Value = FM
if strcmp(app.FiberTreatmentDropDown.Value,'Untreated')
    FTr=[0];
else
    FTr=[1];
end
         
% use a UIFigure and UItable to display data in Apps
fig = app.UIFigure            
%        Aggregating input variables into a table named IP
IP=table(Process,Country,FTr,FWf,FAR,TS,TM,FS,FM,'VariableNames',{'Process','Country','FTr','FWf','FAR','TS','TM','FS','FM'});
UIT = uitable(fig,'Data',IP)

VBBV 2023-4-30

ok, do you still get error ? can you share the code ?

Aditi Mahajan 2023-5-5

在 MATLAB Online 中打开

  
    properties (Access = private)
        Process
        Country
        FTr
        FWf
        FAR
        TS
        TM
        FS
        FM
        IP % Description
    end
    % Callbacks that handle component events
    methods (Access = private)
        % Value changed function: ManufacturingProcessDropDown
        function ManufacturingProcessDropDownValueChanged(app, event)
            app.Process=string(app.ManufacturingProcessDropDown.Value);
        end
        % Value changed function: FiberWeightFractionEditField
        function FiberWeightFractionEditFieldValueChanged(app, event)
            value=app.FiberWeightFractionEditField.Value;
%        Normalizing
            app.FWf=(value-5)/45;
        end
        % Value changed function: FiberAspectRatioEditField
        function FiberAspectRatioEditFieldValueChanged(app, event)
            value=app.FiberAspectRatioEditField.Value;
%        Normalizing
            app.FAR=(value-3.15)/996.85;
        end
        % Value changed function: FiberTreatmentDropDown
        function FiberTreatmentDropDownValueChanged(app, event)
            if strcmp(app.FiberTreatmentDropDown.Value,'Untreated')
                app.FTr=0;
            else
                app.FTr=1;
            end
        end
        % Value changed function: GeographicalLocationDropDown
        function GeographicalLocationDropDownValueChanged(app, event)
            app.Country=app.GeographicalLocationDropDown.Value;
        end
        % Value changed function: TensileStrengthMPaEditField
        function TensileStrengthMPaEditFieldValueChanged(app, event)
            value=app.TensileStrengthMPaEditField.Value;
%        Normalizing
            app.TS=(value-1.62)/49.753;
        end
        % Value changed function: FlexuralStrengthMPaEditField
        function FlexuralStrengthMPaEditFieldValueChanged(app, event)
            value=app.FlexuralStrengthMPaEditField.Value;
%        Normalizing
            app.FS=(value-2.06)/104.34;
        end
        % Value changed function: TensileModulusGPaEditField
        function TensileModulusGPaEditFieldValueChanged(app, event)
            value=app.TensileModulusGPaEditField.Value;
%        Normalizing
            app.TM=(value-0.0834)/4.0006;
        end
        % Value changed function: FlexuralModulusGPaEditField
        function FlexuralModulusGPaEditFieldValueChanged(app, event)
            value=app.FlexuralModulusGPaEditField.Value;
%        Normalizing
            app.FM=(value-0.187)/4.913
        end
        % Button pushed function: CheckforsuitablematrixButton
        function CheckforsuitablematrixButtonPushed(app, event)
            %        Loading classification neural network
            net=load('MSB.mat');
%        Aggregating input variables into a table named IP
            app.IP=table('Size',[1 9], ...
                'VariableTypes',{'categorical','categorical','double','double','double','double','double','double','double'}, ...
                'VariableNames',net.MSB.PredictorNames);
            app.IP.Process(1)=categorical(app.Process);
            app.IP.Country(1)=categorical(app.Country);
            app.IP.FTr(1)=app.FTr;
            app.IP.FWf(1)=app.FWf;
            app.IP.FAR(1)=app.FAR;
            app.IP.TS(1)=app.TS;
            app.IP.TM(1)=app.TM;
            app.IP.FS(1)=app.FS;
            app.IP.FM(1)=app.FM;
%        Predicting the output from the input
            Pred_M=predict(net.MSB,app.IP);
%        Displaying the output 
            app.UsethegivenMatrixEditField.Value=char(Pred_M);