(or at least submitted to CRAN by then)
So - what's new?
I've re-written all the underlying R code, created three new libraries
and updated 2 others
so that all the data processing and manipulation is handled at the C++ level.
The exception to this rewrite is the R API - as a user you should see minimal changes to the code you need to write to ues it (apart from
add_geojson(), I've had to change this slightly). Sorry.
What does this mean for me?
As well as performance improvements, like this example of reducing a 60-second plot to 3 seconds, you can also use one of the new features, such as...
You can make shapes transition smoothly between states
Auto-centering on layer
When you plot data the map will auto-zoom to your layer, so you don't have to specify the
## Old mapdeck( location = c(144.9, -37.8), zoom = 6) %>% add_path( data = roads ) ## new mapdeck() %>% add_path( data = roads )
add_sf() is a convenience function, designed for when you don't really know what's inside your
sf object. It will plot any
sf object by calling the appropriate plotting function on your behalf.
- sf POINTS -
- sf LINESTRINGS -
- sf POLYGONS -
- sf GEOMETRY -
I've given the user more control on how GeoJSON can be plotted, and the function arguments now match the other layers. This is so when using
add_sf( data = sf, fill_colour = "myCol"), the
fill_colour gets propogated to the
add_geojson() layer, etc.
Seems like everything's done then?
My v1.0 list is ever expanding:
and the people working on Deck.gl add features at an alarming rate, so I need to keep up with them too.
Lots more to come!