Basic layout for doing simple cluster analysis in R using the Neotoma Package, across the forest, prairie boundary.

ClusterAnalysisStart.R

library(neotoma)

ostracode <- get_dataset(datasettype='ostracode surface sample', loc=c(-100,43,-92,48))
chemistry <- get_dataset(datasettype='water chemistry',          loc=c(-100,43,-92,48))
diatoms   <- get_dataset(datasettype='diatom surface sample',    loc=c(-100,43,-92,48))
pollen    <- get_dataset(datasettype='pollen surface sample',    loc=c(-100,43,-92,48))

ostracode_dl <- get_download(ostracode)
chemistry_dl <- get_download(chemistry)
diamtom_dl   <- get_download(diatoms)
pollen_dl    <- get_download(pollen)

#  This works because these methods were designed for pollen data largely.
#  Compiles the pollen data to 
pollen_dc <- compile_downloads(compile_taxa(pollen_dl, 'P25'))
pollen_dc_ppt <- pollen_dc[,11:34] / rowSums(pollen_dc[,11:34], na.rm=TRUE)

plot(hclust(dist(pollen_dc_ppt)), cex=0.5)

#  These don't because we have variable column lengths & other issues in the data that need to be addressed in the 'counts' method.
chem_counts      <- do.call(rbind.data.frame,counts(chemistry_dl))
ostracode_counts <- do.call(rbind.data.frame,counts(ostracode_dl))
diatom_counts <- do.call(rbind.data.frame,counts(diatom_dl))