为什么我在进行参数 bootstrap 拟合优度测试时不能超过 nsim=99？

Question

这是我的拟合优度测试：

(fm <- distsamp(~1 ~ndvi_avg, keyfun="hazard", umf))

# Function returning three fit-statistics.

fitstats <- function(fm) {
  
  observed <- getY(fm@data)
  
  expected <- fitted(fm)
  
  resids <- residuals(fm)
  
  sse <- sum(resids^2)
  
  chisq <- sum((observed - expected)^2 / expected)
  
  freeTuke <- sum((sqrt(observed) - sqrt(expected))^2)
  
  out <- c(SSE=sse, Chisq=chisq, freemanTukey=freeTuke)
  
  return(out)
  
}

(pb <- parboot(fm, fitstats, nsim=100, report=1))

我最多可以做 nsim=99，然后我尝试 nsim=100，结果如下：

> (pb <- parboot(fm, fitstats, nsim=100, report=1))
t0 = 103.385 540.1714 70.38591 
Running in parallel on 3 cores. Bootstrapped statistics not reported.

请问这是不是未标记包的bug？

Answer 1

在 parboot() 中尝试 parallel = FALSE。

文档：https://www.rdocumentation.org/packages/unmarked/versions/1.1.0/topics/parboot

为什么我在进行参数 bootstrap 拟合优度测试时不能超过 nsim=99？

Why can I not do more than nsim=99 when doing a parametric bootstrap goodness of fit test?

r

lattice

goodness-of-fit

unmarked-package