R：如何通过R中的最大似然估计得到拟合图像？

Question

这个问题太困扰了我......这是我的R程序：

x<-c(0.06,2.8441,540,1.1,4.974,509.8,66,800.14,3670,762.2336,0.05,2880,2.4514,1235.921,0.57,0.2,270,159.2482,3.6,769,4.65,280,0.1876,0.3,0.07,2200,0.0695,3970,280,138,1050.952,287.2281,5.7128,0.77,2.964,10.2,3.4849,2.7415,1000,1.3796,13.6403,9.3983,322.371,0.8,0.89,110.3466,2,0.38,10,136.0477,0.4,348.8425,148,52,117,240,1350.358)
y=log10(x)
# location and scale based on the y
y.n<-length(y)
y.location<-mean(y)
y.var<-(y.n-1)/y.n*var(y)
y.scale<-sqrt(3*y.var)/pi
# load the stats4 package
library(stats4)
# function to calculate negative log-likelihood
ll.logis <- function(location=y.location, scale=y.scale){
    -sum(dlogis(y, location, scale, log = TRUE))
} 
# fitting logistic distribution using mle function
fit.mle <- mle(ll.logis, method="Nelder-Mead")
# the estimation of parameter
fit.location <- coef(fit.mle)[1]
fit.scale <- coef(fit.mle)[2]

现在，我希望得到这样的Fitting图像，enter image description here 我应该写什么代码？谢谢！

Answer 1

您可以使用ecdf和plot：

plot(ecdf(y))
f <- function(x) plogis(x, fit.location, fit.scale)
plot(f, col="red", add=TRUE, xlim=extendrange(y,f=0.5)

结果如下：