Custom Data Visualization On The Fly

Altair Employee

Oct 2, 2022

Updated Oct 2, 2022 by omet

What if there would be a way to perform quick visualizations using GUI data in your own favourite format and style.

Data Visualization is one of the most important aspects of any design exploration process and visualization comes in various forms such as charts, tables, plots provided by the software being used. Format and style of visualizing data can be subjective, and a design explorer may prefer to perform post-processing in other ways than what is provided by the software. Custom visualization may be just as simple as specifying different fonts, symbols used in plots, colour options in legends and so on. However, often these preferences are broad and are typically expensive to be fulfilled by development teams.

To meet such expectations, Altair HyperStudy offers two options to personalize data visualization: "custom reports" and "utility tool". Both methods are python-based, and this blog will focus on the “utility tool” option.

As mentioned above, "utility tool" is python-based and what makes it attractive over "custom reports" is that it allows selecting data from tables in the GUI. Such flexibility of the tool brings agility in data visualization meaning that variables of interest can be changed without regenerating any kind of report.

The blog will cover a case in which the plotly library from python has been leveraged to create bubble plots. Thickness 1 and Thickness 2 design variables are represented in x and y axis, respectively. And for size and colour dimensions Mass and Displacement responses are chosen, respectively. In Figure 1, the respective columns are chosen and the registered utility function, plotly_bubbleplot is invoked.

Figure 1. Variables are selected from Evaluation Data table.

The utility function renders the bubble plot into a browser which is Chrome in this case as shown in Figure 2. However, if you are not familiar with visualization libraries of Python, data can be rendered into any platform of your choice if the library supports them.