Estimating new data points based on existing spatial observations
Spatial interpolation is a method of estimating new data points based on existing spatial interpolation. This tag focusses on spatial interpolation (data with a geographic component), although interpolation in a general also applies to e.g. timeseries.
Commonly, spatial interpolation involves using of the surrounding observations to estimate a new data point. This can be done using some kind of weighted mean (inverse distance weighted interpolation), by fitting a mathematical function through the existing points (splines), or a combination of both (kriging with external drift).
More information regarding which packages can be used to perform interpolation can be found on the Spatial Task View on CRAN, but in general the gstat
, automap
, Fields
, and geoR
packages are a good start for a wide range of spatial interpolation methods.