On Debial-like distros you have to remove 'libopenblas-base' package. Marcin W dniu 30.12.2015 o 16:22, oleg_komashko(a)ukr.net pisze: -- Marcin Błażejowski GG: 203127

On Thu, 31 Dec 2015, oleg_komashko(a)ukr.net wrote: That's because we were ignoring the potential problem of OpenBLAS running pthreads under OpenMP. Now we're not ignoring it any more. As Jack has said, we don't have firm evidence that this is a practical problem. If it were, the most likely symptom would be a hang in OpenBLAS. At one point lately we were worried that mixing the two methods was leading to a crash, but that turned out not to be the case -- the crash had a different cause, namely an unrelated bug in OpenBLAS, against which gretl now has a work-around. See https://github.com/xianyi/OpenBLAS/issues/723 The basic point, however, is that the interaction of OpenMP and pthreads is undefined. The OpenBLAS library can be built to use OpenMP instead of pthreads, via the flag "USE_OPENMP = 1" in Makefile.rule. Some Linux distros (e.g. Fedora) offer pre-built OpenBLAS in both "flavours": pthreads or OpenMP. Allin

Dear All I have tried my best to find out ARDL estimation using Gretl but I couldn't do so. Is there any method available for this purpose. Please let me know if it could be used for time series analysis. On 30 Dec 2015 20:23, <oleg_komashko(a)ukr.net&gt; wrote:

My package lagreg is pending in the staging area it can do ARDL, find IRF, bootstrap CI and build plots of IRF with CI also, there is more advanced piece  dynMultipl.gfn by Artur Tarassow. My package can be downloaded from the staging area: Warning it's not approved yet, so can have bugs, but IRF was checked against other soft. Oleh P. S. Sometimes ago Allin asked (I myself was among the "convicted") to start a new thread if you have you own topic (which is, obviously the case) 31 грудня 2015, 19:41:06, від "Sven Schreiber" < svetosch(a)gmx.net >: Am 31.12.2015 um 07:35 schrieb Muhammad Akram: Does this question have anything to do with OpenBLAS as in the thread title? It doesn't seem so, so please start a new thread (fresh message to gretl-users) with your problem. That said, ARDL is just OLS with some lags. Use the OLS interface, it comes with lag specification possibilities. cheers, sven _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users

I think it's a good idea You can download it at http://ricardo.ecn.wfu.edu/gretl/staging_fnfiles/lagreg.zip I'm not sure, it seems less general but more easy to use It's intended mostly for intermediate level courses Also I almost  finished functions for automated mean and median lags computing Oleh 1 січня 2016, 13:39:01, від "Artur Tarassow" < artur.tarassow(a)googlemail.com >: Actually the dynMultipl-Package needs some revision. I already started to update the package and hopefully will upload it soon. Btw, Oleh: I've got also some piece of code for a bootstrap version of it. Maybe we could initiate some joint work on it. Actually it would be nice to write an ARDL-toolbox comprising model estimation, ECM-based cointegration tests and bootstrap dynamic multiplier computation... I've already got the pieces but these need to be put together. Happy new year to all! Artur Am 31. Dezember 2015 18:55:23 MEZ, schrieb oleg_komashko(a)ukr.net : My package lagreg is pending in the staging area it can do ARDL, find IRF, bootstrap CI and build plots of IRF with CI also, there is more advanced piece  dynMultipl.gfn by Artur Tarassow. My package can be downloaded from the staging area: Warning it's not approved yet, so can have bugs, but IRF was checked against other soft. Oleh P. S. Sometimes ago Allin asked (I myself was among the "convicted") to start a new thread if you have you own topic (which is, obviously the case) 31 грудня 2015, 19:41:06, від "Sven Schreiber" < svetosch(a)gmx.net >: Am 31.12.2015 um 07:35 schrieb Muhammad Akram: Does this question have anything to do with OpenBLAS as in the thread title? It doesn't seem so, so please start a new thread (fresh message to gretl-users) with your problem. That said, ARDL is just OLS with some lags. Use the OLS interface, it comes with lag specification possibilities. cheers, sven _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users ! ! Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users -- Diese Nachricht wurde von meinem Android-Mobiltelefon mit K-9 Mail gesendet. _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users

Hi Oleh and Jack! Am 02.01.2016 um 17:49 schrieb Riccardo (Jack) Lucchetti: The package looks interesting. I'll have a closer look at it, Oleh! Wow, that's really a clever application of the filter() function! Thank's for this! Artur

On Sat, 2 Jan 2016, oleg_komashko(a)ukr.net wrote: I have no time to play any active role in this, but feel free to ask if you thing I can be of any help on this. Yes, I had noticed that too. However, differences are minor (once you normalise income). The fact that the loglikelihood reported in the book is a little lower makes me think that Greene must have used software (most likely, written by himself --- he's an excellent coder) that uses a slightly more lenient convergence criterion than we do. ------------------------------------------------------- Riccardo (Jack) Lucchetti Dipartimento di Scienze Economiche e Sociali (DiSES) Università Politecnica delle Marche (formerly known as Università di Ancona) r.lucchetti(a)univpm.it http://www2.econ.univpm.it/servizi/hpp/lucchetti -------------------------------------------------------

Thanks! < Greene must have used software (most < likely, written by himself --- he's an excellent coder) I happened to be official translator of Analysis, after that I couldn't look at it for several years! Then I found that since he makes LIMDEP he writes how to program better than most of other authors Oleh