Trend Modelling
The coding of trends, in particular multivariate trends, will be described here.
Martin Schlather, schlather@math.uni-mannheim.de, https://www.wim.uni-mannheim.de/schlather/
data(ca20) ## data set originally from geoR
head(ca20.df)
RFoptions(coordnames=c("east", "north"), varnames="data")
## covariance model with variance, scale and nugget to be estimated;
## just to abbreviate later on
M <- RMexp(var=NA, scale=NA) + RMnugget(var=NA)
## short definition of a trend using the fact that ca20.df is a
## data.frame
ca20.RFmod02 <- ~ 1 + altitude + M
(ca20.fit02.RF <- RFfit(ca20.RFmod02, data=ca20.df, M=M))
## long definition which also allows for more general constructions
ca20.RFmod02 <- NA + NA*RMcovariate(ca20.df$altitude) + M
(ca20.fit02.RF <- RFfit(ca20.RFmod02, data=ca20.df))
## Note that the following also works.
## Here, the covariance model must be the first summand
ca20.RFmod02 <- M + NA + ca20.df$altitude
print(ca20.fit02.RF <- RFfit(ca20.RFmod02, data=ca20.df))
### The following does NOT work, as R assumes (NA + ca20.df$altitude) + M
### In particular, the model definition gives a warning, and the
### RFfit call gives an error:
(ca20.RFmod02 <- NA + ca20.df$altitude + M)
try(ca20.fit02.RF <- RFfit(ca20.RFmod02, data=ca20.df)) ### error ...
## factors:
ca20.RFmod03 <- ~ 1 + area + M ###
(ca20.fit03.RF <- RFfit(ca20.RFmod03, data=ca20.df, M=M))Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.