@kylebarron I went through and renumbered/updated notebooks 04-07 and partially 09 (there's a big # TODO heading at the point where I stopped). Couldn't do 08 cause I can't install Tensorflow on my M1 Mac.
I also used Sidecar to display maps in all notebooks, feel free to revert if you don't want that everywhere. Made the experience of using the notebook a lot better though, will say.
There is one semi large issue: the widget callbacks that we use in Filter and Crossfilter examples to show have updated the slider or selecting from a histogram can change the map are very unstable and cause the kernel to hang. I'm fairly certain this is because of the kernel patching we did to make the sync API work. The reverse works fine: event handlers set on the map to update the histogram/scatterplot seem to function as expected. It's the slider.observe(update_value_filter) and hist_fig.data[0].on_selection(on_histogram_selection_change) etc calls that break.
@kylebarron I went through and renumbered/updated notebooks 04-07 and partially 09 (there's a big # TODO heading at the point where I stopped). Couldn't do 08 cause I can't install Tensorflow on my M1 Mac.
I also used Sidecar to display maps in all notebooks, feel free to revert if you don't want that everywhere. Made the experience of using the notebook a lot better though, will say.
There is one semi large issue: the widget callbacks that we use in Filter and Crossfilter examples to show have updated the slider or selecting from a histogram can change the map are very unstable and cause the kernel to hang. I'm fairly certain this is because of the kernel patching we did to make the sync API work. The reverse works fine: event handlers set on the map to update the histogram/scatterplot seem to function as expected. It's the
slider.observe(update_value_filter)
andhist_fig.data[0].on_selection(on_histogram_selection_change)
etc calls that break.