# Script for DFA from Davis and McHorse, 2013. Palaeo Electronica.
# "A method for improved identification of postrcania from mammalian fossil assemblages: multivariate discriminant function analysis of camelid astragali."

# Genus-level DFA with Jackknife; change first term of the lda command to call another column to run Species or Lumped analysis
# Change CV = FALSE to run analysis without Jackknife.
# Prior is specified to be uninformative; if not included, default prior is set by relative abundances in the training set.
# For data, save the data file (Appendix 1) as a .csv file.

rm()
setwd("Filepath")
camels <- read.csv("data.csv",header=T)
library(MASS)
fit <- lda(Genus..2. ~ LM + TD + TI + TP + LL + WD + WI + LI, data=camels, prior= c(1,1,1,1,1)/5, na.action="na.omit", CV=TRUE)
fit # show results

# Assess the accuracy of the prediction
# percent correct for each category of G
ct <- table(camels$Genus..2., fit$class) 
diag(prop.table(ct, 1))
# total percent correct
sum(diag(prop.table(ct)))
# diagonal table of IDs
ct