Re: [Gretl-devel] A short wish list
gretl-devel-bounces(a)lists.wfu.edu skrev 2013-01-31 21:58:08 :
Från: Allin Cottrell <cottrell(a)wfu.edu>
Till: Gretl development <gretl-devel(a)lists.wfu.edu>
Datum: 2013-01-31 22:03
Ärende: Re: [Gretl-devel] A short wish list
Sänt av: gretl-devel-bounces(a)lists.wfu.edu
On Mon, 28 Jan 2013, andreas.rosenblad(a)ltv.se wrote:
> If possible, I would like to see the following features included in
gretl
> and available form the GUI, to make it more usable:
>
> Univariate tests:
> *Z-test of means for paired data
> *McNemars test
> *Kruskal-Wallis test
> *Friedman's test
> *Jonckheere-Terpstra test
>
> For duration data:
> *Kaplan-Meier curves plot
> *log-rank test
> *Cox regression
>
> Regression models:
> *conditional logistic (logit) regression
> *generalized estimating equations
> *generalized linear mixed models
>
> Any chances of any of these being included?
The chances are good if anyone is willing to write function
packages to handle these things. The current gretl development
model is that we're not just going to code a whole bunch of
extra things in C (unless they're arguably essential and
are much faster if written in C). Users who have mastered the
basics of hansl are invited to contribute extra tests and so
on.
See http://users.wfu.edu/cottrell/tmp/extending.pdf . Actually
I need to update this; there's now some extra functionality to
make for better integration of contributed packages into the
GUI.
Allin Cottrell
I would argue that at least all tests and methods used in introductory
courses in statistics for business and economics are essential and should
be included in the default/basic gretl. This would make gretl even more
useable for such courses. Currently missing are at least Z-test of means
for paired data and Kruskal-Wallis test.
Also, for duration data, I think that Kaplan-Meier curves plot, log-rank
test and Cox regression are basic and essential tests and methods.
Andreas Rosenblad