[ENH] plotting for libpysal raster functionality

[slumnitz]

space to discuss visualisations for new raster functionality

Idea to include plotting functionality for weights generated using raster data see

This would include:

suggestions:

• to refactor plot_spatial_weights() to automatically detect and plot a raster weights object
• potentially to 'zoom' new plot_spatial_weights() plots to areas with no data values, to check whether weights where build correctly, or create new functionality called check_spatial_weights() that does the according including plotting a raster multiplot (to be discussed)
• plot small multiples and weights only?
• in general to refactor plot_spatial_weights() to use the new planned id_order() attribute.
• how to efficiently build weights lines? possibility to plot weights lines on raster instead of creating polygons?

Examples can be seen in @MgeeeeK raster examples notebook

space for declined aspects:

• [ ]

[MgeeeeK]

I'd a small doubt related to the first point... gdf argument of plot_spatial_weights() method accepts only geopandas dataframe, since we're going to make it work with both raster and vector data, therefore we need another argument to pass DataArray but this wont be similar to automatically detecting the type of the data.

[MgeeeeK]

I was also thinking of combining weights plot along with the raster plot (like a two column subplots), will add a link to a notebook for showing the prototype.