# Indicator Spcies Analysis を簡単に実行する関数
isa <- function(df, row.data=F){	# $stand, $species, $cover, $community
	table <- tapply(df$cover, list(df$stand, df$species), sum)	# make compositional table
	table[is.na(table)] <- 0
	table <- data.frame(table)	#  to be coerced to data.frame
	name.table <- colnames(table)	# original names of species
	colnames(table) <- as.numeric(factor(colnames(table)))	# numerical names of species
	names.table <- data.frame(species=name.table, numeric.sp=colnames(table))
	com <- unique(subset(df, select=c(stand, community)))$community	# unique stand and community
	com <- as.numeric(com)	# to be coerced to numeric
	res <- duleg(table, com)	# Species Indicator Analysis
	if(!row.data){
		res <- data.frame(numeric.sp=names(res$maxcls), community=res$maxcls, ind.val=round(res$indcls,3), p.value=round(res$pval,4))
		res <- merge(names.table, res)
		res <- res[order(res$community, res$ind.val, res$p.value, decreasing=T),][,-1]
	}
	return(res)
}

# 使用方法
library(akima)	# load library
library(labdsv)	# load library

# データ整理
data(bryceveg)	# returns a vegetation dataframe
df <- as.data.frame.table(as.matrix(bryceveg))
df <- subset(df, Freq!=0)
colnames(df) <- c("stand","species","cover")
df <- merge(df, data.frame(stand=unique(df$stand), community=sample(1:5, length(unique(df$stand)), replace=TRUE)))

isa(df)	# 整理した解析結果を出力
isa(df, row.data=T)	# 生の解析結果を出力

df.2 <- read.table( "http://kobe.cool.ne.jp/matsut/df.txt", header=T)	# 別のデータ
isa(df.2)	# 整理した解析結果を出力
isa(df.2, row.data=T)	# 生の解析結果を出力