R Programming/Method of Moments

      • Package gmm implements the generalized method of moment and the generalized empirical likelihood.

      First, it is possible to estimate a simple linear model or a simple linear model with instrumental variables using the gmm() function. The GMM method is often used to estimate heteroskedastic instrumental variable models.

      > # Simple linear model
      > N <- 1000
      > u <- rnorm(N)
      > x <- 1 + rnorm(N)
      > y <- 1 + x + u
      > res <- gmm(y ~ x, x)
       
      > # Simple linear model with instrumental variables.
      > library(gmm)
      > N <- 1000
      > u <- rnorm(N)
      > z <- rnorm(N)
      > x <- 1 + z + u + rnorm(N)
      > y <- 1 + x + u
      > res <- gmm(y ~ x, z)
      > summary(res)
      
      Last modified on 29 January 2013, at 10:18