search

predict-idlab / plotly-resampler

Visualize large time series data with plotly.py
https://predict-idlab.github.io/plotly-resampler/latest
MIT License
990 stars 67 forks source link

[FEAT] improve documentation of the aggregators #276

Closed jonasvdd closed 6 months ago

jonasvdd commented 7 months ago

The documentation on how plotly-resampler works; i.e. the time-series data aggregation should be more clear (and less hidden).

jonasvdd commented 7 months ago

Will add this in the main Readme; updating the docs will happen when we omit the TraceUpdater dependency and rely solely on Patch

plotly-resampler is a powerful enhancement for standard Plotly figures, designed to significantly improve the visualization of line charts. It achieves this by implementing dynamic data aggregation, which adjusts based on the current view in the front-end. Essentially, plotly-resampler enriches Plotly figures with the capability to handle large datasets more efficiently through this dynamic aggregation approach.

This functionality is primarily driven by advanced time-series data point selection algorithms. For these algorithms, plotly-resampler utilizes highly optimized implementations provided by tsdownsample.

The package's default aggregation method is MinMaxLTTB (with 1000 selected data points). For a deeper understanding of this method, you can refer to its dedicated repository and the associated research paper, which offer comprehensive details and insights into its workings and advantages.