setPoints(), setFeatures() and setGroups() so that new tables can only be set via joining by the primary key (fid, fid and gid respectively).gtTheme now contains a separate slot for all information that relate to the colour scale of values and the legend.getLayer() also for sf and Spatial objects.visualise()This update introduces the geom type ‘grid’, which is meant to store raster data. It stores as @point slot a compact version that gives (in both x and y dimension) the three values origin, number of cells and cell distance/resolution. It stores values in run-length encoding (rle) (if that is smaller than non-compressed). The @feature slot of a grid geom can contain any number of tables that would correspond to layers in a raster. It can store in the @group slot an attribute table per each of the layers.
The following functions have been modified:
gc_geom() now has a method to convert Raster objects to grid geoms.get/setTable(x, slot = "point") has been moved to get/setPoints(x)get/setTable(x, slot = "feat") has been moved to get/setFeatures(x)get/setTable(x, slot = "group") has been moved to get/setGroups(x)getPoints() for grid geoms extracts the compacted @points slot into the interoperable table this getter is supposed to hand out.getFeatures() for grid geoms reconstructs fid for points and extracts rle tables, if they exist.makeObject() is not a generic anymore but is now a function based on getters. That means that any class that has the respective getters defined, can be plotted with visualise() (which includes now sf and sp objects).getSubset() has been integrated into the respective getters, i.e., to get a subset of points, use getPoints(x, fid == 1), etc.The following functions are new:
gc_raster() to transform a grid geom to an object of class Raster*.getRes() to get the resolution of a gridded object.getLayer() to extract layers from an object that has layers.gs_sketch().