Is it possible to fit a truncated Poisson or Negative Binomial regression model in R? If so, how? Thanks.