Hansel: A Deducer Plug-In for Econometrics

This paper discusses the development of a Deducer plug-in, referred here to as Hansel, that can deal with techniques typically found in undergraduate courses in econometrics, along with some more advanced econometric techniques (the final package name should be something like DeducerHansel). Currently the Deducer package (Fellows, 2012) provides an exceptional interface that deals with a number of areas including generalized linear models. Thus it can already deal with ordinary least squares, weighted least squares, probit models and logit models. However it is not currently wellsuited for dealing with time-series data, panel data, or censored data, or for dealing with instrumental variables. That is where Hansel helps. The following areas are among those covered by Hansel: twostage least squares; tobit models; smoothing, filtering, and forecasting; unit root testing; vector autoregressive models; cointegration testing; and various panel data techniques. Hansel can deal with the time series classes ts, zoo, and xts in addition to data frames. Hansel is similar in ease to the commercial software EViews and another open-source econometric software package called gretl, which is written in C. Hansel is not only useful for students in econometrics courses, but also provides an opportunity for those unacquainted with R to quickly get down to the business of using it for estimation. This can provide a gateway for deeper use of R.