Small function to create PCA analysis and spatial eigenvariables

eigenvariables.fct.R

eigenvariables.fct <- function(vars, name, proportion = 0.95){
  library("raster")
  if (file.exists("./pca") == FALSE) dir.create("./pca")
  #Running PCA:
  # Counts not NA cels
  non.na <- sum(!is.na(values(vars[[1]])))
  # Sample the study area with n-non.na and creates an environmental table
  sr <- sampleRandom(vars, non.na)
  # faz o PCA dessa tabela padronizada
  pca <- prcomp(scale(sr))
  summary.pca <- summary(pca)
  #Saving results:
  capture.output(pca, file = sprintf('./pca/%s.pca.txt', name))

  #saving summary
  capture.output(summary.pca, file = sprintf('./pca/%s.summary.pca.txt', name))

  #Plotting results
  #GGPLOT
  #####
  #library(ggplot2)
  # create data frame with scores
  #scores  <-  as.data.frame(pca$x)
  # plot of observations
  #ggplot(data = scores, aes(x = PC1, y = PC2, label = rownames(scores))) +
  #  geom_hline(yintercept = 0, colour = "gray65") +
  #  geom_vline(xintercept = 0, colour = "gray65") +
  #  geom_text(colour = "tomato", alpha = 0.8, size = 4) +
  #  ggtitle("PCA plot of USA States - Crime Rates")
  #####
  # png(filename = sprintf('./pca/%s.pca.biplot.png',name),
  #     bg = "white")
  # biplot(pca)
  # dev.off()

  # Creating eigenvariable in space
  axis.nb <- which(summary.pca$importance["Cumulative Proportion",] >= proportion)[1]
  eigenvariables <- predict(vars, pca, index = 1:axis.nb)

  if (file.exists("./env") == FALSE) dir.create("./env")
  writeRaster(eigenvariables,sprintf('./env/%s.eigenvariables.tif',name),overwrite=T)
}

Inserida variavel 'proportion', para definirmos qual proporcao de explicaçao (logo, quantos eixos iremos salvar) da PCA. proportion=entre 0 e 1