从ictreg模型自动计算BIC

我从

列表

包提供的

ictreg（）

估计的模型中寻找一种计算贝叶斯信息标准（BIC）的方法，但没有找到答案

例如：

library(list)

data(race)

lm.results <- ictreg(y ~ south + age + male + college, data = race, 
                 treat = "treat", J=3, method = "ml")

summary(lm.results)

---

这是我遇到的解决方案。它使用模型使用以下公式计算BIC：

-2*对数似然+npar*对数（nobs）

，其中npar表示参数数量，nobs表示拟合模型中的观测数量

以下函数计算BIC：

bic_ictreg <- function(fit) {
  # Check if method is "ml". Stop if not.
  if(fit[["method"]] != "ml") {
    stop("Function needs a model that uses the Expectation-Maximization algorithm to estimate. Use method = 'ml' within ictreg")
  }

  # Calculate BIC
  -2*fit[["llik"]] + (length(fit[["par.treat"]]) + 
                        length(fit[["par.control"]])) * 
    log(length(fit[["pred.post"]]))
}

bic_ictreg(lm.results)

结果与Blair&Imai报告的BIC一致（2012年，第70页）

参考文献

Blair，G.，和Imai，K.（2012）。列表实验的统计分析。政治分析，20（01），47-77

bic_ictreg <- function(fit) {
  # Check if method is "ml". Stop if not.
  if(fit[["method"]] != "ml") {
    stop("Function needs a model that uses the Expectation-Maximization algorithm to estimate. Use method = 'ml' within ictreg")
  }

  # Calculate BIC
  -2*fit[["llik"]] + (length(fit[["par.treat"]]) + 
                        length(fit[["par.control"]])) * 
    log(length(fit[["pred.post"]]))
}

bic_ictreg(lm.results)

library(list)

data(race)

lm.results <- ictreg(y ~ south + age + male + college, data = race, 
                 treat = "treat", J=3, method = "ml")

bic_ictreg(lm.results)

[1] 2959.797