library(MASS)
Iris <- data.frame(rbind(iris3[,,1], iris3[,,2], iris3[,,3]),
Sp = rep(c("s","c","v"), rep(50,3)))
train <- sample(1:150, 75)
z_1 <- lda(Sp ~ ., Iris, prior = c(1,1,1)/3, subset = train)
pre_1 <- predict(z, Iris[-train, ])
z_2 <- qda(Sp ~ ., Iris, prior = c(1,1,1)/3, subset = train)
pre_2 <- predict(z, Iris[-train, ])
Fisher判别分析,即 LDA相应的R实现为:
MASS包中的 lad() 函数,qda() 函数
lad(x, grouping, prior = proportions ,tol = 1.0e-4, method , CV = FALSE, nu, .....)
lad(formula, data, .... ,subset , na.action )
GGA是广义密度近似,LDA是局域密度近似。对于计算能量,LDA计算所得值一般大于GGA。对于弱相互作用,以前有人说LDA好,现在看来是GGA要好一些。对于不同体系还要看传统选择,参考文献。