Overview

Polyaxon comes with a powerful and customizable visualization system for driving insights in the dashboard or programmatically.

Oftentimes users have advance use cases that are not supported by the platform, or they need to integrate visualizations generated by external systems.

Polyaxon can consume any chart specification generated by these libraries:

  • Matplotlib
  • Plotly
  • Bokeh
  • Altair/Vega

Matplotlib

You can log charts generated by Matplotlib directly to Polyaxon.

Polyaxon supports two methods for logging matplotlib figures:

  • Logging figures as images
  • Parsing the figures to interactive charts

Matplotlib static figures

from polyaxon import tracking

figure = ...

tracking.init(...)

tracking.log_mpl_image(figure, name="my-figure", step=12)

run-dashboards-matplotlib-static

More info about tracking api

Matplotlib interactive figures

from polyaxon import tracking

figure = ...

tracking.init(...)

tracking.log_mpl_plotly_chart(figure=figure, name="my-figure", step=12)

run-dashboards-matplotlib-interactive

More info about tracking api

Plotly

You can log charts generated by Plotly directly to Polyaxon.

from polyaxon import tracking

figure = ...

tracking.init(...)

tracking.log_plotly_chart(figure=figure, name="my-figure", step=12)

run-dashboards-plotly

More info about tracking api

Bokeh

You can log charts generated by Bokeh directly to Polyaxon.

from polyaxon import tracking

figure = ...

tracking.init(...)

tracking.log_bokeh_chart(figure=figure, name="my-figure", step=12)

run-dashboards-bokeh

More info about tracking api

Altair

You can log charts generated by Altair or vega-light directly to Polyaxon

from polyaxon import tracking

figure = ...

tracking.init(...)

tracking.log_altair_chart(figure=figure, name="my-figure", step=12)

run-dashboards-altair

More info about tracking api