Dashboard (UI) software reference¶
Overview¶
As previously mentioned in the the Basic Usage section, the dashboard-style user interface for the system is built using a widget based approach, that relies on the Jupyter interactive computing platform.
This approach was chosen for the following reasons:
As a Python-based project, the Jupyter platform is a natural fit, providing a popular, well-documented Python environment that is capable of producing rich interactive, web-based visualisations, using python libraries and code which should be familiar to undergraduates who have a grounding in Python programming.
Modern web browsers provide an ideal deployment platform for graphical applications, offering compatibility across a wide range of devices, operating systems and user interfaces (mouse, or touchscreen).
Structure¶
The underlying Python code for the dashboard application is composed of two main types of files:
Widget definition files, largely composed of ipywidget object declarations, which define the appearance of the user interface, in addition to some local behaviour, specific to the user elements defined within a given file.
Utility function files, containing regular python code that defines useful functions/classes that are then imported in widget definition files and used by the widget objects in them, to effect global behaviour of the user interface, such as handling of shared data and progression of the application state.
A ‘top-level’ widget definition file, app.py, is then used to bring together the various UI elements defined in the
other files, in a nesting/container-like manner, producing the final dashboard interface.
This approach introduces a great deal of flexibility, when it comes to implementing additional UI elements and behaviours, as it is only necessary to create a new widget definition file and then import that into the top-level script.
The present structure of the UI code includes the following files: