Final Words

Again, before finishing the module I would like to give you some insights into the main concepts you should have become familiar with.

1. You must be able to access spatial layers, both vector and raster.

2. Be aware we can use vector layers as a source of information of point-like events to later interact with raster data.

3. RasterStacks can be used as source data for predict() so that we can produce a spatial prediction from any regression model .

4. The validation procedure based on random samples remains the same. We do not use spatial prediction on them since it does not make any sense.