[R] ​ use of PPML (Poisson Pseudo Maximum Likelihood)

Achim Zeileis Achim.Zeileis at uibk.ac.at
Thu Oct 8 18:55:32 CEST 2015


On Thu, 8 Oct 2015, ?????? ?. wrote:

> Dear Colleagues!
>
>
> We use R for research. We would like to perform calculations (gravity model
> of trade) with the
> ??
> use of PPML (Poisson Pseudo Maximum Likelihood).
>
> Please explain:
>
> 1. What package in R enables us to make calculations in accordance with the
> PPML?
>
> We used the command
> glm()

Yes.

> with the parameter
>
> family=quasipoisson(link="log").

That is one option. In econometrics, the quasi-Poisson version is less 
commonly used than in statistics, though. Alternatively, so-called 
"robust" sandwich standard errors can be applied, i.e.,

m <- glm(..., family = poisson)

library("lmtest")
library("sandwich")
coeftest(m, vcov = sandwich)

> Does this command procedure PPML? If not, then what command allows us to
> use PPML?
>
> 2. For calculations using a linear model the quality factor is R^2. What 
> is the analogue of this indicator for PPML in system R?

That is not really an R question. Sometimes people use pseudo-R-squared 
measures, e.g., available in package "pscl". Whether or not this is really 
useful is a different question, though.

> Thank you very much in advance!
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.



More information about the R-help mailing list