Sorry, can't remember who mentioned this, but someone said it would be nice to see the numerical values of the roots in this plot. This is done in CVS and will be in the next release: when you mouse over the "dot" representing a root you get a label showing the numerical value.

From: Allin Cottrell <cottrell(a)wfu.edu&gt; Reply-To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; Subject: Re: [Gretl-users] VAR roots plot Date: Sun, 22 Jan 2006 17:50:46 -0500 (EST) On Sun, 22 Jan 2006, Sven Schreiber wrote: >Allin Cottrell schrieb: >>Sorry, can't remember who mentioned this, but someone said it would be >>nice to see the numerical values of the roots in this plot. >> >>This is done in CVS and will be in the next release: when you mouse over >>the "dot" representing a root you get a label showing the numerical >>value. >> > >I'm not the one who asked for this, but if it is included (which is >great btw) I suggest to add the possibility of displaying the full list >of roots. Otherwise you need to write down the values by hand. That's a reasonable suggestion. Should be doable. Allin Cottrell _______________________________________________ Gretl-users mailing list Gretl-users(a)ricardo.ecn.wfu.edu http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users

From: jack <jack(a)deanovell.unian.it&gt; Reply-To: jack(a)deanovell.unian.it, Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; Subject: Re: [Gretl-users] VAR roots plot Date: Mon, 23 Jan 2006 10:28:51 +0100 (CET) On Mon, 23 Jan 2006, john w wrote: > Hi, > I mentioned VAR root plots. > Ok. Thank you. And I have another question/request. > What about Information criteria (AIC, BIC, HQ etc.) for selecting lags in > VAR/VECM? > LR test for the longest lag (lag signifficance) is great but it would be > nice to have Information criteria tests for choosen number of lags in one > table. > > What do you think? This is rather easy to do. However, a dedicated function does the job nicely too. See the attached files. Riccardo `Jack' Lucchetti Dipartimento di Economia Università Politecnica delle Marche jack(a)dea.unian.it http://www.econ.univpm.it/lucchetti

<< var_autolag.inp >>

<< var_autolag_ex.inp >>

_______________________________________________ Gretl-users mailing list Gretl-users(a)ricardo.ecn.wfu.edu http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users

From: jack <jack(a)deanovell.unian.it&gt; Reply-To: jack(a)deanovell.unian.it, Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; Subject: Re: [Gretl-users] VAR roots plot Date: Mon, 23 Jan 2006 11:36:29 +0100 (CET) On Mon, 23 Jan 2006, john w wrote: > Thnx Jack. > 1. This could be done for VECM too? Yes, of course, although normally you select the lag length first on the VAR representation and only then you work on the cointegration rank. But again, you can write a similar function for vecms rather easily. > 2. function var_autolag(list X, list Z, scalar maxlag). X,Z and maxlag must > be changed manually? > If there are no exogenous then Z must be 0? The function is meant to be used from within a script. I thought the file var_autolag_ex was self-explanatory (but obviously wasn't :-)) You can either write a script or run the following commands in a gretl console: 1) run "include var_autolag.inp" to load the function in memory 2) define the lists with any name you want; for example list foo = x1 x2 list bar = z1 z2 with the proviso that you can use the syntax list bar = null for an empty list. 3) run var_autolag(foo,bar,4) or autolags = var_autolag(foo,bar,4) if you want to save the results in a vector (may be useful for futher processing) > 3. In your previous mails you mentioned GUI. Where I can download GUI (if > you have some link)? Oh, sorry, GUI stands for "Graphical User Interface". What I meant is "what you get by running the graphical version of gretl rather than the command-line version". Riccardo `Jack' Lucchetti Dipartimento di Economia Università Politecnica delle Marche jack(a)dea.unian.it http://www.econ.univpm.it/lucchetti _______________________________________________ Gretl-users mailing list Gretl-users(a)ricardo.ecn.wfu.edu http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users

From: jack <jack(a)deanovell.unian.it&gt; Reply-To: jack(a)deanovell.unian.it, Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; Subject: Re: [Gretl-users] VAR roots plot Date: Mon, 23 Jan 2006 13:05:35 +0100 (CET) On Mon, 23 Jan 2006, john w wrote: > Thank you Jack. I'll try. > Regarding my point 3. In Gretl VAR module (GUI) deterministic variables > covers trend and exogenous variables? > This is from Gretl help: ...any "deterministic" or exogenous terms ... > > In VECM module the terminology is quite clear, but in VAR module a little > bit cunfuses me. > My suggestion for VAR module is: > 1. To include one more check option for including a trend (like for a > constant), The way the source code is now, that would be a little inconvenient, unless we're willing to add a "--trend" option to rhe "var" command. I wouldn't be against it, but adding a trend to a VAR is also very easy now, and such an option would be a tad gratuitous. Hence, I'm not either strongly pro nor against such a change. What do other list members think? > 2. To replace "Deterministic variables" with "Exogenous variables". Any adjective you use in this context is rather arbitrary. The word "exogenous" may mean at least 4 different things in the econometrics literature, so my very personal preference is for using it as sparingly as possible, unless the context makes what I mean very clear. But others may disagree. Riccardo `Jack' Lucchetti Dipartimento di Economia Università Politecnica delle Marche jack(a)dea.unian.it http://www.econ.univpm.it/lucchetti _______________________________________________ Gretl-users mailing list Gretl-users(a)ricardo.ecn.wfu.edu http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users

From: Sven Schreiber <svetosch(a)gmx.net&gt; Reply-To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; Subject: Re: [Gretl-users] VAR roots plot Date: Mon, 23 Jan 2006 13:30:53 +0100 john w schrieb: > You are wrigt Jack. > At the end, it's better to work on completely new features rather than > spending time on cosmetics. Let it stay. > don't give up so quickly ;-) >> > >> > In VECM module the terminology is quite clear, but in VAR module a >> little >> > bit cunfuses me. >> > My suggestion for VAR module is: >> > 1. To include one more check option for including a trend (like for a >> > constant), >> >> The way the source code is now, that would be a little inconvenient, >> unless we're willing to add a "--trend" option to rhe "var" command. I >> wouldn't be against it, but adding a trend to a VAR is also very easy >> now, >> and such an option would be a tad gratuitous. Hence, I'm not either >> strongly pro nor against such a change. What do other list members think? I would (mildly) favor consistency for the interface of VARs and VECMs, but it's not terribly important. >> >> > 2. To replace "Deterministic variables" with "Exogenous variables". >> >> Any adjective you use in this context is rather arbitrary. The word >> "exogenous" may mean at least 4 different things in the econometrics >> literature, so my very personal preference is for using it as >> sparingly as >> possible, unless the context makes what I mean very clear. But others may >> disagree. In this context "deterministic" seems wrong, because you can add any variable you like. "Exogenous" may not be precise, but at least it's correct. So I think john w is right and it should be changed (that way it's also consistent with the vecm formulation). cheers, sven _______________________________________________ Gretl-users mailing list Gretl-users(a)ricardo.ecn.wfu.edu http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users

From: Sven Schreiber <svetosch(a)gmx.net&gt; Reply-To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; To: Gretl list <gretl-users(a)ricardo.ecn.wfu.edu&gt; Subject: Re: [Gretl-users] VAR roots plot Date: Mon, 23 Jan 2006 14:04:09 +0100 john w schrieb: > Ah ah my friend. Never! :))) > Look how many posts are mine. > > Always forward! > But as I see they are only two (Allin and Jack). Let's insist on new > features firstly. > Well, changing one word to another is not going to prevent them from adding new features, I'd say. And I think a good and consistent user interface is very important especially for a program like gretl. In terms of raw econometric and number-crunching features there are quite a few alternatives out there. What sets gretl apart (imho) is mainly (but not only, of course) the intuitive interface. -sven _______________________________________________ Gretl-users mailing list Gretl-users(a)ricardo.ecn.wfu.edu http://ricardo.ecn.wfu.edu/mailman/listinfo/gretl-users