Statistical data visualization dashboard
A crucial aspect of the statistical data analysis is the visualization of the processed data. We want to offer the possibility for download (see #129) and via scripts (see #276), but it should be possible to display some basic visualizations in the Web UI of the model data explorer.
The workflow for this shall be developed within this user story. We aim for generating a dashboard where the user can access the analysis requests (see #255) and add them to a graph via drag and drop. When inspecting the detail page of an analysis request (i.e. the page where the user is redirected to, when he submits the request), one will also see such a dashboard, but here, the corresponding request is already added to the graph.
We can also implement the possibility to dynamically add new graphs to this dashboard, and we can implement the functionality to choose different plot methods, e.g. filled area plots for uncertainty intervals (standard deviation, percentiles, etc.), bar plots for precipitation. Another option would be the interpolation of two statistical graphs for a comparison via metrics such as R2.
The potential for these comparisons are very rich but we have to make sure, that we limit the methodologies (for now) to a basic, easy-to-understand and simple visualization. We can add more functionalities at a later state.
The scope of this user story is to define
- what we want
- and how to design the dashboard (in the prototype)
As usual, the development team will come up with an initial approach for the implementation, discuss it with the team, revise it and discuss it again.
related User Story (Hereon internal access only): #283