home.social

#holoviz β€” Public Fediverse posts

Live and recent posts from across the Fediverse tagged #holoviz, aggregated by home.social.

fetched live
  1. There is always so much more to tell. I only realized afterwards that I overlooked the new spatial and temporal unit support 🀦

    But I did remember the new and improved support and our plans for / integration

  2. Exploring the data app publishing options of .

    panel convert (panel.holoviz.org/how_to/wasm/) looks super promising but tricky since both and seem to be missing from

    But for now, enjoy this demo viewer for visualizing data from @datagvat

    mybinder.org/v2/gh/anitagraser

  3. I'm so happy that we're finally in a place where we can whip up a quick trajectory data exploration app with maps and graphs from rather arbitrary inputs, be it plain old csv with x/y in any crs known to proj, or fancy gis formats in just a few minutes.

    It's minimal extra effort and seems to impress most people so much more than comparable plots in a notebook πŸ€·β€β™€οΈπŸ˜€

  4. Heard a nice talk today in #Berlin from Jean-Luc Stevens about #Holoviz at @pyberlin . #hvPlot and #Datashader look like a powerful combination to render large datasets. Jean-Luc works like @sophiamyang for #Anaconda
    holoviz.org/index.html

  5. I love the idea of turning notebooks into slides. I've figured out most issues with rendering interactive plots πŸ‘©β€πŸ’» but there is one thing I'm stuck with:

    & works great in a but is not not rendered in the slides

    The output is: BokehModel(combine_events=True, render_bundle={'docs_json': {'8bb4dc3b-92b9-4d6e-9b03-060b443bfa47': {'defs': …

    Any pointers much appreciated

  6. Today's success moment: came up with a hack to have arrow markers in πŸ‘©β€πŸ’» while building a app with @geopandas & .

    I'm still surprised that the built-in @bokeh markers don't include any arrows. None of the available markers can be used to unambiguously show direction docs.bokeh.org/en/latest/docs/

  7. Started with #holoviz #panel today. Complex python code wrapped up into a nice dashboard, including #ipyleaflet, which you can quickly pack into a #docker image and run somewhere like fly.io. The last bit isn't quite yet working, but very happy with how things have turned up!