How can I access Ydata from interaction plots?

Hi,

I am studying interaction effects of independent variables on dependent variables in Matlab using interaction plots. I have several interaction plots like this:

I want to collect the Y-values of all the plots. Therefore, I would like to create a code which can access the Ydata from each plot.

I know that [h,AX,bigax] = interactionplot(...) allows us to access the subplot axes data, but I do not know how to extract the data using these handles. Any suggestions?

[Adrsch3p]

Instead of acessing the data from interactionplot, I would suggest that you get it directly from the function grpstats. Because Y data of the interactionplot is in fact computed using grpstats.

Answer 1

[Star Strider]

Using an example from the interactionplot documentation —

rng default;     % For reproducibility
y = randn(1000,1);

% Randomly generate data for four three-level factors.

group = ceil(3*rand(1000,4));

% Display the interaction plots for the factors and name the factors 'A', 'B', 'C', 'D'.

[h,AX,bigax] = interactionplot(y,group,'varnames',{'A','B','C','D'});

Axx = findobj(h.Children, 'Type','Axes');
for k1 = 1:size(Axx,1)-1
    Axk = Axx(k1);
    Lines = findobj(Axk.Children,'Type','Line');
    for k2 = 1:numel(Lines)
        xc{k1,k2} = Lines(k2).XData;
        yc{k1,k2} = Lines(k2).YData;
        nc{k1,k2} = Lines(k2).DisplayName;
    end
    tm = [xc(k1,:);yc(k1,:);nc(k1,:)];
    C(k1,:) = tm(:).';
end

% xc{:}                                               % View Results (Optional)
% yc{:}                                               % View Results (Optional)
% nc{:}                                               % View Results (Optional)

DNC = [compose('X%d',1:3); compose('Y%d',1:3); compose('DisplayName%d',1:3)];
Results = cell2table(C, 'VariableNames',DNC(:).')

Results = 16×9 table
        X1                           Y1                      DisplayName1        X2                         Y2                     DisplayName2        X3                           Y3                      DisplayName3
    ___________    ______________________________________    ____________    ___________    ___________________________________    ____________    ___________    ______________________________________    ____________

    1    2    3      -0.10012      -0.14598       0.16843     {'A = 3' }     1    2    3      0.05729     -0.13914    -0.061251     {'A = 2' }     1    2    3     -0.082502      0.061337     0.0041465     {'A = 1' } 
    1    2    3      -0.10012      -0.14598       0.16843     {0×0 char}     1    2    3      0.05729     -0.13914    -0.061251     {0×0 char}     1    2    3     -0.082502      0.061337     0.0041465     {0×0 char} 
    1    2    3      -0.11877     -0.091477        0.0791     {0×0 char}     1    2    3    -0.098697    -0.093319     0.031328     {0×0 char}     1    2    3     -0.035499      0.011612    0.00083478     {0×0 char} 
    1    2    3      0.046984     -0.052158      -0.10695     {0×0 char}     1    2    3    -0.040539     -0.16683      0.04365     {0×0 char}     1    2    3     -0.044517      -0.13393       0.16396     {0×0 char} 
    1    2    3     0.0041465     -0.061251       0.16843     {0×0 char}     1    2    3     0.061337     -0.13914     -0.14598     {0×0 char}     1    2    3     -0.082502       0.05729      -0.10012     {0×0 char} 
    1    2    3    8.8326e-05      0.021481      0.071188     {'B = 3' }     1    2    3     -0.24391     -0.18516      0.11863     {'B = 2' }     1    2    3    -0.0070887     -0.020265      -0.09547     {'B = 1' } 
    1    2    3    8.8326e-05      0.021481      0.071188     {0×0 char}     1    2    3     -0.24391     -0.18516      0.11863     {0×0 char}     1    2    3    -0.0070887     -0.020265      -0.09547     {0×0 char} 
    1    2    3       0.14227      -0.10678      0.057161     {0×0 char}     1    2    3     -0.17583     -0.15826     0.044711     {0×0 char}     1    2    3     -0.018925     -0.084299     -0.015907     {0×0 char} 
    1    2    3    0.00083478      0.031328        0.0791     {0×0 char}     1    2    3     0.011612    -0.093319    -0.091477     {0×0 char}     1    2    3     -0.035499     -0.098697      -0.11877     {0×0 char} 
    1    2    3      -0.09547       0.11863      0.071188     {0×0 char}     1    2    3    -0.020265     -0.18516     0.021481     {0×0 char}     1    2    3    -0.0070887      -0.24391    8.8326e-05     {0×0 char} 
    1    2    3     -0.010002     -0.035674       0.17971     {'C = 3' }     1    2    3     0.020492     -0.13946     -0.04292     {'C = 2' }     1    2    3     -0.044294      -0.19478     -0.041124     {'C = 1' } 
    1    2    3     -0.010002     -0.035674       0.17971     {0×0 char}     1    2    3     0.020492     -0.13946     -0.04292     {0×0 char}     1    2    3     -0.044294      -0.19478     -0.041124     {0×0 char} 
    1    2    3       0.16396       0.04365      -0.10695     {0×0 char}     1    2    3     -0.13393     -0.16683    -0.052158     {0×0 char}     1    2    3     -0.044517     -0.040539      0.046984     {0×0 char} 
    1    2    3     -0.015907      0.044711      0.057161     {0×0 char}     1    2    3    -0.084299     -0.15826     -0.10678     {0×0 char}     1    2    3     -0.018925      -0.17583       0.14227     {0×0 char} 
    1    2    3     -0.041124      -0.04292       0.17971     {0×0 char}     1    2    3     -0.19478     -0.13946    -0.035674     {0×0 char}     1    2    3     -0.044294      0.020492     -0.010002     {0×0 char} 
    1    2    3       0.16396       0.04365      -0.10695     {'D = 3' }     1    2    3     -0.13393     -0.16683    -0.052158     {'D = 2' }     1    2    3     -0.044517     -0.040539      0.046984     {'D = 1' } 

This returns cell arrays of the x-data (‘xc’) , y-data (‘yc’), and ‘DisplayName’ (‘nc’) for each line in the nested plots. Address them by their row and column numbers. To find out what are in the various cells, use the ‘nc’ information and then use those row-column indices to find the corresponding ‘xc’ and ‘yc’ information.

The table isn’t perfect, since the first line of every set is duplicated, however it should be enough to get the information yuou want.

I use the example from the documentation to test the code, and since I have never previously used interactionplot, I have no experience with it. I have no idea how fragile the code may be to other data sets.

.