Skip to content

Point Cloud Visualization

Jump to bottom
Michael Dyck edited this page Sep 17, 2020 · 2 revisions

Point Cloud Visualization

Finally, this component visualizes the results of the previously simulated Unity executable in an interactive Python GUI. It does so by reading the Results.txt file created by the Evaluation Framework's second component.

  • Activate your Python environment as described here
  • Navigate to the VisualizationGUI folder of the repository
  • Execute the following command (NameOfTheEvaluationToVisualize is the previously specified name when defining the evaluation space, in this case DemonstrationEvaluation):
python point_cloud.py [NameOfTheEvaluationToVisualize]
  • optionally, you can specify a floating point value for the time scale at which simulations in Unity will be presented (see below), e.g. --timeScale 0.5 for half execution speed
  • As a result, the Python GUI window opens:
  • The whole window can be resized as needed to fit any screen sizes, the fraction of the upper and lower part of the window can be resized individually

Upper Frame

  • The upper frame of the GUI visualizes the results of the evaluation, where each point represents one combination of EvaluationParamters in the EvaluationSpace
  • For both x- and y-axis, as well as the color bar, the user can freely choose which of the EvaluationParameters or ObjectiveValues the respective axis represents, making it possible to investigate the results from all possible perspectives and compare different ObjectiveValues
  • Additionally, the user can extend the visualization to show a 3D plot of the results, adding a z-axis
  • The toolbar in the lower left corner of the upper frame allows for dragging, zooming and saving as it is common in plots generated by Python's Matplotlib

Lower Frame

  • The lower frame depicts a list of all EvaluationParameters and ObjectiveValues present in the current evaluation
  • The two fields next to the sliders show the current minimum and maximum value of the respective parameter
  • The sliders allow the user to "deselect" specific paramters in the plot to focus on special parts of the Evaluation Space by dragging either the left or right cursor
  • Alternatively, the user can manually enter a number into one of the two fields besides the sliders to limit the range of depicted values
  • The red values on the right hand side of the lower frame depict the values of all EvaluationParameters and ObjectiveValues of the currently selected "solution"
    • The user can select a solution simply by clicking on a point in the plot, which automatically gets circumvented by a black circle
    • Clicking the Launch Current Solution button now launches the Unity executable of the evaluated project and simulates the scene visible for the user for exactly the currently selected set of EvaluationParamters
    • This simulation is now executed at the speed previously defined through the --timeScale 0.5 command line parameter when launching the GUI (default: 1, normal speed)

Feel free to play around with the GUI and its interactive elements as you wish and explore, how the GUI can be used to gain a deep understanding of the Evaluation Space and the influence of different EvaluationParameters on the ObjectiveValues

Back to the Demonstration Project

© Siemens AG, 2020

Author: Michael Dyck (m.dyck@gmx.net)

Clone this wiki locally