April 2016 - Gretl-users - gretlml.univpm.it

by Jan Tille

Dear all, I tried to estimate a random effects probit model using the --robust option. After this, I re-estimated the model without it. I got the same results (i.e. standard errors) in both cases and looking at the output it became clear why, because Gretl said that estimation was "...based on inverse Hessian". However, I found this post from 2013, where Sven Schreiber already commented on this issue. http://lists.wfu.edu/pipermail/gretl-devel/2013-December/004835.html However, I did not find an answer to this post, which brings me to my question, is it possible use robust standard errors when one estimates a random effects probit model in Gretl. By "robust" I mean robust to some form of unspecified misspecification, like one checks the robust option, when using ols. Kind regards, Jan

8 years, 8 months

3
5
0 / 0

Granger Causality via menu?

by Wingenroth, Thorsten

Hi, is it possible to call the Granger Causality test (GC) via the menu? I tried with Model/Time Series/Vector autoregression but there are no lags for the exogenous variable (and they cannot be specified, the button is gray). Calling GC in Hansl with the var command works perfectly in contrast. In short: Is this a bug or a feature ;-) ? Thanks and kind regards, Thorsten

8 years, 8 months

3
4
0 / 0

2 graphing issues

by Summers, Peter

Hi all, I've just noticed a couple of things when plotting a series: 1. The "recession bars" option seems to have disappeared (this is a monthly time series data set) 2. If I add a fitted line to the plot (eg, cubic), click OK, then re-open the edit menu and switch back to "none" (no fitted line), I get what seems to be a spurious error message. A box pops up saying "'*' - no numeric conversion performed!", but when I click close, the fitted line disappears as it should. This is Gretl 2016b, build date 4-14, on Windows 7. TIA, PS [cid:logoimg]

8 years, 8 months

2
3
0 / 0

Non linear Granger causality tests

by Ioannis A. Venetis, Assoc. Prof.

Dear Jose, In Panchenko's site http://research.economics.unsw.edu.au/vpanchenko/ [4] you will find http://research.economics.unsw.edu.au/vpanchenko/software/2006_GC_JEDC_c_... [5] at paper 15. You can run the code (exe file) in windows (works for me). Ioannis A. Venetis, Associate Professor Director of Postgraduate Studies for the M.Sc "Applied Economics & Data Analysis" University of Patras, School of Business Administration, Department of Economics University Campus, Rio, 26504, Greece Tel: 0030 2610 969964 http://www.econ.upatras.gr/en/people [1] [2]https://sites.google.com/site/ioannisavenetis/ [3]************************************************************************************************** Links: ------ [1] http://www.econ.upatras.gr/en/people [2] http://application.econ.upatras.gr/en/node/130 [3] https://sites.google.com/site/ioannisavenetis/ [4] http://research.economics.unsw.edu.au/vpanchenko/ [5] http://research.economics.unsw.edu.au/vpanchenko/software/2006_GC_JEDC_c_...

8 years, 8 months

2
1
0 / 0

Non linear Granger causality tests

by JOSE FRANCISCO PERLES RIBES

Dear list: First of all excuse me because it is not exactly an specific Gretl issue. But, I'm trying for several days to find some code in Gretl, Eviews or other software (R,Gauss, etc.) to perform the Hiemstra and Jones (1994) or the Dicks and Pachenko (2006) non-linear causality test and I can't find it. Somebody on the list has a code or knows any software that performs these tests?. Thanks in advance and sorry for any inconvenience. José Perles University of Alicante.

8 years, 8 months

2
1
0 / 0

Install Gretl-2016b in Debian Jessie

by Carlos Andrade

Hello everybody, How to install Gretl-2016b in Debian Jessie? There is a backport? Thanks, -- Atenciosamente, Prof. Carlos A. S. de Andrade LAPEA - Laboratório de Pesquisa em Economia Aplicada e Engenharia de Produção Universidade Federal de Campina Grande. Centro de Humanidades Unidade Acadêmica de Economia

8 years, 8 months

1
0
0 / 0

issue with fctstats()

by Artur T.

Dear all, I am not sure this behavior is intended, but to me it looks some unintended behavior. In the following example one of the scalars in the vector of realized values y is zero which drives the division "(e./y)=-inf" and which affects the later sum calculation as well. I guess gretl should ignore this "-inf"-value, or do I miss anything here? <hansl> matrix y = {0; 1} matrix f = {0.5;2} matrix e = y-f eval fcstats(y,f) # MPE by hand eval e./y mat = (0.5).*sumc(100*e./y ) mat </hansl> Best, Artur

8 years, 8 months

2
6
0 / 0

gretl 2016b released

by Allin Cottrell

For the benefit of anyone not on the gretl-announce list: there's a new release on sourceforge today. Changelog entry: 2016-04-14 version 2016b - "panel" command, random-effects version: show an overall Wald chi-square test on the named regressors; also support the --robust option - "dataset" command: expose and document the "pad-daily" option - "vecm" command: clarify the options available when specify- ing a restriction on beta via matrices R and q - Commands that produce both textual output and plots: add a --plot option to control the format of the plot - GUI help text: make references to the Gretl User's Guide chapter-specific, and make the file open at the specified chapter if the PDF viewer supports that functionality - GUI "view code" option for function packages: avoid slow loading of big code buffers - GUI plot editor: enable selection of point size for graph styles that use points - Internal code: introduce several optimizations to speed up execution of complex scripts - "foreign" language mechanism: add basic support for Julia (including editing and execution of Julia programs) - build: try harder to respect --disable-nls (though this is very difficult nowadays) - Fix bug: crash on performing some matrix transformations using the GUI "icon view" - Fix bug: broken format in output of actual versus forecast values in some cases - Fix bug: CSV import failing in locales that use decimal comma when file uses decimal comma and also '.' as thousands marker - Fix bug: handling of "missingness" in imported daily data not working quite right - Fix bug: "omit" command with --auto option failing for ordered logit/probit models - Fix bug: potential crash in printing the results of a random-effects probit model with small RE variance - Fix bug: GUI access to nonparametric models was broken - Fix bug: not flagging an error when given a broken right- hand side specification in the "restrict" command - Fix bug: matrix-form restrictions on VECM model not being accepted even when valid, in some cases - Fix bug: partial parsing of command line when blocked by a false if-condition could give rise to spurious errors - Fix bugs: some subtle fail conditions for saving compiled loops onto functions - OS X (quartz) build: update to GTK 2.24.29, and fix a GTK issue which led to crashes on OS X 10.11 (El Capitan) - OS X (quartz) build: fix problem with gnuplot wxt window, add Adwaita theme, and make Adwaita the default - Bundled gnuplot for Windows and OS X: update to 5.1 cvs as of March 2016 - MS Windows builds: update to OpenBLAS 0.2.18dev. Allin

8 years, 8 months

1
0
0 / 0

string substitution with sprintf?

by Summers, Peter

Hi all, I'm trying to write a script that will produce a LaTeX table as output - basically a customized version of tabprint. Here's the basic idea: <hansl> matrix doh = mnormal(1,3) # this would be a coefficient vector, e.g. string fmt = "%.4g " string baz = "doh[1]" loop i=2..cols(doh) # build up the format and argument strings fmt ~= " & %.4g " baz ~= ", doh[$i]" endloop fmt baz sprintf Foo "\"%s\"\n", fmt Foo # this printf is what I want, but I want to use the "fmt" string above printf "%.4g & %.4g & %.4g\n", @baz # this seems like it should work, but I get a "syntax error" sprintf bar Foo, @baz </hansl> I'm still trying to get my head around (s)printf, string substitution, etc., so I'm probably missing something straightforward. Any suggestions welcome! Thanks, PS

8 years, 8 months

5
7
0 / 0

Re: [Gretl-users] Connectiviy with R

by George Matysiak

Thanks Allin. Also, noted what you say about the 'crash'. George *Contact number +48 781 415 432**"In God we trust, all others bring data."*

8 years, 8 months

1
0
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Gretl-users April 2016