Re: [R] Calculating AIC for the whole model in VAR
On 21/05/2013 16:00, Dimitri Liakhovitski wrote: Hello! I am using package VAR. What is that? There is no such package on CRAN nor BioC. I've fitted my model: mymodel-VAR(mydata,myp,type=const) I can extract the Log Liklihood for THE WHOLE MODEL: logLik(mymodel) How could I calculate (other than manually) the corresponding Akaike Information Criterion (AIC)? I tried AIC - but it does not take mymodel: AIC(mymodel) # numeric(0) Thank you! This is not reproducible, pace the posting guide. The default method for AIC() should work if the logLik() method is written correctly, so I guess it was not. -- Brian D. Ripley, rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Calculating AIC for the whole model in VAR
Sorry, I am using package vars On Tue, May 21, 2013 at 11:09 AM, Prof Brian Ripley rip...@stats.ox.ac.ukwrote: On 21/05/2013 16:00, Dimitri Liakhovitski wrote: Hello! I am using package VAR. What is that? There is no such package on CRAN nor BioC. I've fitted my model: mymodel-VAR(mydata,myp,type=**const) I can extract the Log Liklihood for THE WHOLE MODEL: logLik(mymodel) How could I calculate (other than manually) the corresponding Akaike Information Criterion (AIC)? I tried AIC - but it does not take mymodel: AIC(mymodel) # numeric(0) Thank you! This is not reproducible, pace the posting guide. The default method for AIC() should work if the logLik() method is written correctly, so I guess it was not. -- Brian D. Ripley, rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~**ripley/http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 -- Dimitri Liakhovitski [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Calculating AIC for the whole model in VAR
On 21/05/2013 16:11, Dimitri Liakhovitski wrote: Sorry, I am using package vars So you need to report the bug in that package to its maintainer. ?logLik says Value: Returns an object of class ‘logLik’. This is a number with at least one attribute, ‘df’ (*d*egrees of *f*reedom), giving the number of (estimated) parameters in the model. and the methods in vars do not comply. On Tue, May 21, 2013 at 11:09 AM, Prof Brian Ripley rip...@stats.ox.ac.uk mailto:rip...@stats.ox.ac.uk wrote: On 21/05/2013 16:00, Dimitri Liakhovitski wrote: Hello! I am using package VAR. What is that? There is no such package on CRAN nor BioC. I've fitted my model: mymodel-VAR(mydata,myp,type=__const) I can extract the Log Liklihood for THE WHOLE MODEL: logLik(mymodel) How could I calculate (other than manually) the corresponding Akaike Information Criterion (AIC)? I tried AIC - but it does not take mymodel: AIC(mymodel) # numeric(0) Thank you! This is not reproducible, pace the posting guide. The default method for AIC() should work if the logLik() method is written correctly, so I guess it was not. -- Brian D. Ripley, rip...@stats.ox.ac.uk mailto:rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~__ripley/ http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 tel:%2B44%201865%20272861 (self) 1 South Parks Road, +44 1865 272866 tel:%2B44%201865%20272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 tel:%2B44%201865%20272595 -- Dimitri Liakhovitski -- Brian D. Ripley, rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Calculating AIC for the whole model in VAR
Here is a reproducible example: library(vars) data(Canada) mymodel - VAR(Canada, p = 2, type = const) myLL-logLik(mymodel) AIC(myLL) Why does logLik(mymodel) say that df=NULL? Might this be the reason for AIC(myLL) being numeric(0)? Dimitri On Tue, May 21, 2013 at 11:17 AM, Prof Brian Ripley rip...@stats.ox.ac.ukwrote: On 21/05/2013 16:11, Dimitri Liakhovitski wrote: Sorry, I am using package vars So you need to report the bug in that package to its maintainer. ?logLik says Value: Returns an object of class logLik. This is a number with at least one attribute, df (*d*egrees of *f*reedom), giving the number of (estimated) parameters in the model. and the methods in vars do not comply. On Tue, May 21, 2013 at 11:09 AM, Prof Brian Ripley rip...@stats.ox.ac.uk mailto:rip...@stats.ox.ac.uk** wrote: On 21/05/2013 16:00, Dimitri Liakhovitski wrote: Hello! I am using package VAR. What is that? There is no such package on CRAN nor BioC. I've fitted my model: mymodel-VAR(mydata,myp,type=**__const) I can extract the Log Liklihood for THE WHOLE MODEL: logLik(mymodel) How could I calculate (other than manually) the corresponding Akaike Information Criterion (AIC)? I tried AIC - but it does not take mymodel: AIC(mymodel) # numeric(0) Thank you! This is not reproducible, pace the posting guide. The default method for AIC() should work if the logLik() method is written correctly, so I guess it was not. -- Brian D. Ripley, rip...@stats.ox.ac.uk mailto:rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~__**ripley/http://www.stats.ox.ac.uk/~__ripley/ http://www.stats.ox.ac.uk/~**ripley/http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 tel:%2B44%201865%20272861 (self) 1 South Parks Road, +44 1865 272866 tel:%2B44%201865%20272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 tel:%2B44%201865%20272595 -- Dimitri Liakhovitski -- Brian D. Ripley, rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~**ripley/http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 -- Dimitri Liakhovitski [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Calculating AIC for the whole model in VAR
Am Dienstag, den 21.05.2013, 16:17 +0100 schrieb Prof Brian Ripley: On 21/05/2013 16:11, Dimitri Liakhovitski wrote: Sorry, I am using package vars So you need to report the bug in that package to its maintainer. ?logLik says Value: Returns an object of class ‘logLik’. This is a number with at least one attribute, ‘df’ (*d*egrees of *f*reedom), giving the number of (estimated) parameters in the model. and the methods in vars do not comply. Dear Prof. Ripley, many thanks for pointing this out. The attributes 'df' and 'nobs' have been added to logLik.varest() on R-Forge (project 'AICTS II', revision = 90); soon to be released on CRAN (package version 1.5-1). Best, Bernhard On Tue, May 21, 2013 at 11:09 AM, Prof Brian Ripley rip...@stats.ox.ac.uk mailto:rip...@stats.ox.ac.uk wrote: On 21/05/2013 16:00, Dimitri Liakhovitski wrote: Hello! I am using package VAR. What is that? There is no such package on CRAN nor BioC. I've fitted my model: mymodel-VAR(mydata,myp,type=__const) I can extract the Log Liklihood for THE WHOLE MODEL: logLik(mymodel) How could I calculate (other than manually) the corresponding Akaike Information Criterion (AIC)? I tried AIC - but it does not take mymodel: AIC(mymodel) # numeric(0) Thank you! This is not reproducible, pace the posting guide. The default method for AIC() should work if the logLik() method is written correctly, so I guess it was not. -- Brian D. Ripley, rip...@stats.ox.ac.uk mailto:rip...@stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~__ripley/ http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 tel:%2B44%201865%20272861 (self) 1 South Parks Road, +44 1865 272866 tel:%2B44%201865%20272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 tel:%2B44%201865%20272595 -- Dimitri Liakhovitski __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.