Recent changes to feature-requestshttps://sourceforge.net/p/gretl/feature-requests/2016-11-10T13:04:55Z#97 cross-sectional dependence test2016-11-10T13:04:55Z2016-11-10T13:04:55ZKTTKhttps://sourceforge.net/u/helix1234/https://sourceforge.net600d070fe6752e6cdf76619648c3d949bab5a1c6<div class="markdown_content"><p>As I wrote, not sure about what is used in the wild and if the statistics (their properties) differ substantially. The user-written Stata module xtcsd offers Friedman's test and Frees' test besids Pesaran's CD test:<br/>
<a href="http://fmwww.bc.edu/RePEc/bocode/x/xtcsd.html" rel="nofollow">http://fmwww.bc.edu/RePEc/bocode/x/xtcsd.html</a></p></div>#94 Multilevel regression2016-11-08T13:02:14Z2016-11-08T13:02:14ZSven S.https://sourceforge.net/u/svetosch/https://sourceforge.net6783895ab02b47435f7c1614cb7c41567455e340<div class="markdown_content"><ul>
<li><strong>status</strong>: open --> closed</li>
</ul></div>#94 Multilevel regression2016-11-08T13:02:03Z2016-11-08T13:02:03ZSven S.https://sourceforge.net/u/svetosch/https://sourceforge.netc492ffed1d19d61781133ebc41a3f5268d00813c<div class="markdown_content"><p>The OP contacted me (off-tracker by mistake, I guess) after the previous post, pointing to entire textbooks like "Multilevel and Longitudinal Modeling Using Stata, Volumes I and II". I replied to this:</p>
<p>"Sorry, this is way too generic. As I said, much of what's in these books is perfectly possible already with gretl. You really have to be more specific, otherwise our answer would simply be: The feature is already there.</p>
<p>(Note that the jargon used for the exact same model in various sub-branches of statistics often differs. So maybe that could be a source of a misunderstanding. For example, in the gretl guide you probably don't find much related to the term "multilevel". But that doesn't mean the model isn't there. But that problem is not gretl-specific; we can try to point you in the needed directions, but only if the problem definition is more specific.)"</p>
<p>So I'm closing this request. The offer to "point you in the needed directions" still stands, however.<br/>
cheers,<br/>
sven</p></div>#98 2SLS summary: use chisq test for model2016-11-08T12:56:32Z2016-11-08T12:56:32ZSven S.https://sourceforge.net/u/svetosch/https://sourceforge.netd75d753e309c0654de3350d36c1bbfbde366447e<div class="markdown_content"><ul>
<li><strong>status</strong>: open --> closed</li>
</ul></div>#98 2SLS summary: use chisq test for model2016-11-08T12:56:13Z2016-11-08T12:56:13ZSven S.https://sourceforge.net/u/svetosch/https://sourceforge.net184439be4a38e3fb6b01d69ac17f2aa92ec7a24c<div class="markdown_content"><p>First, I would really recommend to re-join the gretl (users') mailing list, where such discussions are very welcome.<br/>
So, since no feature is requested, I will close this.<br/>
However, the question as such is valid; my answer would be in principle, yes, similar to that in the other ticket. (Asymptotically it's all the same anyway.) I'm currently looking at eq. (8.55) in Davidson&MacKinnon for example, where an F statistic for the IV case is constructed via artificial regressions. So the only question is, which d.o.f. does gretl use here? (I haven't looked at the actual output right now, probably it's given in the output.)</p></div>2SLS summary: use chisq test for model2016-11-07T19:19:26Z2016-11-07T19:19:26ZKTTKhttps://sourceforge.net/u/helix1234/https://sourceforge.netea727e8714973e491754197fa0b1e5bc61784dd2<div class="markdown_content"><p>I noticed a slight difference between the summary statistics for 2SLS between Stata and Gretl:<br/>
Gretl uses a F test while Stata uses a chisq test<br/>
(see e.g., <a href="https://www.stata.com/manuals13/rivregress.pdf" rel="nofollow">https://www.stata.com/manuals13/rivregress.pdf</a> , example 1, p. 6)<br/>
coefficient tests seems to be carried out by both programmes using the normal distribution (z values).</p>
<p>Note: This might be related to ticket #93 (random effects model: F statistic or wald chisq statistic ))</p>
<p>Any reference about the "correct" test and the distribution? Same argument as in #93?</p></div>2SLS summary: use chisq test for model2016-11-07T19:19:26Z2016-11-07T19:19:26ZKTTKhttps://sourceforge.net/u/helix1234/https://sourceforge.net46f1258d74a97567292f823829c077b2d6beebcb<div class="markdown_content"><p>Ticket 98 has been modified: 2SLS summary: use chisq test for model<br/>
Edited By: Sven S. (svetosch)<br/>
Status updated: u'open' => u'closed'</p></div>#97 cross-sectional dependence test2016-11-02T09:21:44Z2016-11-02T09:21:44ZKTTKhttps://sourceforge.net/u/helix1234/https://sourceforge.net2e3d8ff0c4cd219a4e74f7b19e19a932c1fba0d8<div class="markdown_content"><p>This is great news, thank you guys!</p>
<p>How do you think about implementing the LM <span>[1]</span> and scaled LM <span>[2]</span> test for cross sectional dependence as well? Should be pretty easy to compute those once the components for Pesaran's CD statistic are there. However, I am not sure if those are used widly in applied econometrics...</p>
<p><span>[1]</span> Breusch, T.S. and A.R. Pagan (1980), The Lagrange multiplier test and its applications to model specification in econometrics, Review of Economic Studies, 47(1), pp. 239–253.</p>
<p><span>[2]</span> Pesaran, H. (2004), General Diagnostic Tests for Cross Section Dependence in Panels, CESifo Working Paper 1229.</p></div>#64 dashed and dotted lines in time series plots2016-11-01T10:20:11Z2016-11-01T10:20:11ZSven S.https://sourceforge.net/u/svetosch/https://sourceforge.netdf2784b1a6d743df287347f991c591a58057a066<div class="markdown_content"><p>Interesting new developments: See this thread <a href="http://lists.wfu.edu/pipermail/gretl-users/2016-October/012121.html" rel="nofollow">http://lists.wfu.edu/pipermail/gretl-users/2016-October/012121.html</a> for a hint about a solution. I'm repeating my own example here:<br/>
<hansl><br/>
open denmark.gdt --quiet<br/>
list Lplot = LRM LRY<br/>
plot Lplot<br/>
options with-lines time-series<br/>
literal set linetype 1 dashtype 2<br/>
literal set linetype 2 dashtype 2<br/>
# literal set mono<br/>
end plot --output=display<br/>
</hansl> </p>
<p>What remains to be done is to enhance gretl's plotting GUI with the corresponding dashtype choosers. (But of course that would finally mean requiring gnuplot version 5 also on older Linux distros, if I understand the issue correctly.)</p></div>#94 Multilevel regression2016-10-29T16:01:14Z2016-10-29T16:01:14ZSven S.https://sourceforge.net/u/svetosch/https://sourceforge.net58f44db0b0e1ec1380e478c047a2a9f407c59db7<div class="markdown_content"><p>Again, please specify what exactly do you have in mind that's missing from gretl. Some stuff for example could be handled by interaction effects, see for example the paragraph "Generating lists of transformed variables" in chapter 14.1 of the user guide. Or paragraph "Interaction dummies" in chapter 16.3.<br/>
If we don't hear from you or somebody else with similar interests until the end of the year this will be closed.</p></div>