Skip to content
Mauro Masiero edited this page Feb 22, 2023 · 6 revisions

Guide on best practices with plotly

Stacked barplot

Available at pgx.stackedBarplot. Documentation coming soon.

Boxplot

Available at pgx.boxplot.PLOTLY. Documentation coming soon.

Barplot (option with or without mean +/- sd)

Available at pgx.barplot.PLOTLY. Documentation coming soon.

Plotting large datasets:

Plotly worked well with large datasets, both across tabs, multiple plots, and ~1M points. To reduce rendering time and memory usage, possible solutions:

  • plotly marker has max display, which limits number of plots displayed. Needs further investigation on backend algo.
  • for plots that do not require all points (eg. volcano): sort the points based on their importance, eg. test statistics and give those points priority. Inform users that this is happening.
  • Datashader: currently only available in Python.

Creating subplots (grids) with plotly:

Plotly has different options to combine multiple plots. Those can be focused on different aspects of the app interface. Some solutions are here:

Clone this wiki locally