Question

我的概率分布函数是

f(x;r)=(1+r)/Γ(1/(1+r))exp(-x^(1+r)) , 0<x<Inf, r>-1

我正在使用以下代码：

library(stats4)   
x3=c(0.927869895,0.000559193,0.059761785,0.361542986,0.274291999,0.563373589,
     0.565878385,0.126281994,0.46623362,0.796111638,0.809460469,0.28011156,
     0.263443012,0.704551045,0.978333171,0.139962088,0.599336759,0.108009066,
     0.202133384,0.05401611)

nLL <- function(r)-sum(log(1+r)-log(gamma(1/(1+r)))-x3^(1+r))
fit0 <- mle(nLL, start = list(r = 0.1), nobs = NROW(x3))
fit1 <- mle(nLL, start = list(r = 0.1), nobs = NROW(x3),
            method = "Brent", lower = -1, upper = 150)
stopifnot(nobs(fit0) == length(x3))

现在我想提取存储在fit1中的MLE的值。我使用了以下命令

fitv = fit1$Coefficients

但我无法获得该值，因为它给出了以下错误

fit1 $ Coefficients错误：未为此S4类定义$运算符

现在我得到了答案

nLL <- function(r)-sum(log(1+r)-log(gamma(1/(1+r)))-x1^(1+r))
fit0 <- mle(nLL, start = list(r = 0.1), nobs = NROW(x1))
fit1 <- mle(nLL, start = list(r = 0.1), nobs = NROW(x1),
        method = "Brent", lower = -1, upper = 200)
stopifnot(nobs(fit0) == length(x1))
class(fit1)
str(fit1)
fiteg[k] = fit1@coef

如何使用mle函数提取r中的MLE值

0 个答案: