Skip to content

Visualizations / Plots

hetida designer supports visualizations and plots via

  • Plotly (PLOTLYJSON type)
  • arbitrary HTML and images via the ANY type.

Plotly

Via the PLOTLYJSON type for transformation outputs Plotly figures can be used for visualizing data. The test execution result view and the experimental dashboarding render such plots directly.

To develop your own custom plot components we recommend to start with one of the existing plot base components like "Single Timeseries Plot" from the "Visualization" category and make a copy of it. The code of "Single Timeseries Plot" in particular explains how to take into account target timezone and locale information provided by an external system as context to the execution.

Operators with just plot output(s) are only run if the execution runs pure plot operators (run_pure_plot_operators) and their plots are not discarded, e.g. wired to the drop adapter. Executions from the hetida designer user interface run them, executions via the web service endpoints by default do not. Operators providing plots together with other outputs are always run, but their component code can skip creating plots that are not needed. See Skipping Unneeded Outputs for details.

ANY type for arbitrary html

Via ANY type outputs you can provide arbitrary HTML output that is then rerendered by the test execution result view and the experimental dashboarding. An example is

{
  "content_type": "text/html",
  "encoding": "plain",
  "name": "HTML Output via ANY",
  "data": "<h2>HTML example</h2><br>So<b>me</b><br>HTML"
}

See the ANY type documentation for the correct json structure that must be provided for this feature.

This allows to write components that use pure javascript visualization libraries like apache echarts or µplot.

µplot component example

The "µplot Single Timeseries Plot" base component demonstrates how one can integrate µplot to render high-volume timeseries data efficiently.