Main Content

Export Plots in Classification Learner App

After you create plots interactively in the Classification Learner app, you can export your app plots to MATLAB® figures. You can then copy, save, or customize the new figures. Choose among the available plots: scatter plot, parallel coordinates plot, confusion matrix, ROC curve, minimum classification error plot, LIME explanations plot, Shapley explanations plot, and partial dependence plot.

  • Before exporting a plot, make sure the plot in the app displays the same data that you want in the new figure.

  • On the Learn, Test, or Explain tab, in the Export section, click Export Plot to Figure. The app creates a figure from the selected plot.

    • The new figure might not have the same interactivity options as the plot in the Classification Learner app. For example, data tips for the exported scatter plot show only X,Y values for the selected point, not the detailed information displayed in the app.

    • Additionally, the figure might have a different axes toolbar than the one in the app plot. For plots in Classification Learner, an axes toolbar appears above the top right of the plot. The buttons available on the toolbar depend on the contents of the plot. The toolbar can include buttons to export the plot as an image, add data tips, pan or zoom the data, and restore the view.

  • Copy, save, or customize the new figure, which is displayed in the figure window.

    • To copy the figure, select Edit > Copy Figure. For more information, see Copy Figure to Clipboard from Edit Menu.

    • To save the figure, select File > Save As. Alternatively, you can follow the workflow described in Customize Figure Before Saving.

    • To customize the figure, click the Edit Plot button on the figure toolbar. Right-click the section of the plot that you want to edit. You can change the listed properties, which might include Color, Font, Line Style, and other properties. Or, you can use the Property Inspector to change the figure properties.

As an example, export a scatter plot in the app to a figure, customize the figure, and save the modified figure.

  1. In the MATLAB Command Window, read the sample file fisheriris.csv into a table.

    fishertable = readtable("fisheriris.csv");
  2. Click the Apps tab.

  3. In the Apps section, click the arrow to open the gallery. Under Machine Learning and Deep Learning, click Classification Learner.

  4. On the Learn tab, in the File section, click New Session .

  5. In the New Session from Workspace dialog box, select the table fishertable from the Data Set Variable list.

  6. Click Start Session. Classification Learner creates a scatter plot of the data by default.

  7. Change the predictors in the scatter plot to PetalLength and PetalWidth.

    Scatter plot in the app for the Fisher iris data with the predictors PetalLength and PetalWidth

  8. On the Learn tab, in the Export section, click Export Plot to Figure.

  9. In the new figure, click the Edit Plot button on the figure toolbar. Right-click the points in the plot corresponding to the versicolor irises. In the context menu, select Color.

    Exported scatter plot for the Fisher iris data with points selected, and the Color option selected in the context menu

  10. In the Color dialog box, select a new color and click OK.

    Color dialog box

  11. To save the figure, select File > Save As. Specify the saved file location, name, and type.

    Exported plot figure window with the Save As option selected in the File menu

Related Topics