integrating custom deck.gl layer with kepler.gl - what's the best practice?
dear kepler.gl devs, I would like to know, what is the best approach to integrate a custom deckl.gl layer with kepler.gl? (i.e., so that I can select this layer when I choose the data mapping in kepler.)
to give you an example of what I want to achieve: I extended the standard ArcLayer to show an animation indicating the direction of the link - and now I would like to add that as an option to kepler.gl. I followed the tutorials in the vis-academy and checked the documentations and issues of both kepler and deck but I could not find any documentation on how to approach this.
I would greatly appreciate it if you could add a description for how to achieve this, for instance as an additional exercise in the vis academy's kepler.gl tutorial.
From what I gathered so far, the actual layer is defined by deck.gl -- and if it should be usable in kepler.gl, an interface has to be defined, specifying the fields & options that would be visible in the user interface when parametrizing the layer.
however, I'm not sure how to integrate my "updated"/custom version of deck with kepler. and I also do not know how to write the interface on the kepler side, in order to enable parametrization of the new layer in my custom deck.gl version.
please let me know if I should clarify any part of my request. kind regards!
Any further to this? I'd love to use kepler's data management and easy layering to create a story map but I need Trips layer.
I am also interested in using custom deck.gl layer (e.g., pydeck-earthengine-layers) with kepler.
@giswqs It is a nice idea. It is not too hard to add wrapper code for a deck.gl layer, however there is a complication with the EarthEngineLayer in that it requires the application to be logged in to an earth engine enabled google account, so Google Login would need to be added to kepler.gl, which would require some UI design work and some plumbing.
Also the data objects being visualized are not just tables but ee API objects so there would need to be a way to create those.
But I assume your interest is around the jupyter integration of kepler rather than the app, the same problems need to be solved but in a different way, more along the way pydeck-ee did it which requires a some relatively specialized integration work.
Summary: not a low hanging fruit but definitely possible...
@ibgreen Thanks for the insights. Any yes, my interest is around kepler-jupyter rather than the app. Kepler is so powerful for visualizing vector data. It would be nice to integrate pydeck-ee with kepler to visualize the massive raster datasets provided by Earth Engine.
Any update on this ? I'd also love some guidance to integrate custom deck.gl layers (like delaunay interpolation layer) to kepler.gl.