This is a consolidated response to some points made lately by Sven about various aspects of graphing/plotting in gretl. 1) There's now a "bars" mode for the --band-style option to "gnuplot' and "plot":

2) Up till now, the GUI "savename" apparatus for graphs, as in

both save the plot as a named icon and display it in a window

3) This is a bit more technical. In http://lists.wfu.edu/pipermail/gretl-devel/2016-October/006960.html Sven suggests that we move to require gnuplot 5 for gretl (our current minimum supported gnuplot version is 4.6.0). His reasoning is, "I would like to see gretl handing the data over to gnuplot in the 'multiple-columns-side-by-side' format that gnuplot 5 apparently adopted following Allin's suggestion. It would then be much easier to tweak the appearance of plots beyond the subset functionality that is directly supported by gretl." There are various good reasons for requiring gnuplot 5 but I'm not sure this is one of them.

then referenced by "plot". This means the data block must precede "plot". No big deal, I suppose, if the data are few, but with a dense plot you might have to scroll down through hundreds of lines of data to get to the "plot" clauses (e.g. when editing the plot commands in a gretl window). It seems to me this would be more awkward.

But with respect to a longer discussion of HP in the guide I'm not so sure, this could get quite fundamental. For example, here are some other recent interesting (and critical) papers: - Phillips, Peter C.B., and Sainan Jin. 2015. “Business Cycles, Trend Elimination, and the HP Filter,” Cowles Foundation Discussion Paper No. 2005, Yale University. - James D. Hamilton 2016. "Why You Should Never Use the Hodrick-Prescott Filter", http://econweb.ucsd.edu/~jhamilto/hp.pdf But I'm not at all against "aligning" the default smoothing values based on some common criterion (where the criterion value would be arbitrary I guess and would simply imply the convention of 1440 for quarterly data). cheers, sven

Regarding my previous reply. The moment I sent it I realized my wording was rather rude. My apologies. What I meant to say was that I appreciate Ignacio bringing those papers to my attention and that I support a brief discussion be included in docs. I did not mean to imply that it was my decision in any way. I am very happy to leave leading in the very capable hands of people like Allin and Jack. I hope I didn't offend anyone.

Regarding my previous reply. The moment I sent it I realized my wording was rather rude. My apologies. What I meant to say was that I appreciate Ignacio bringing those papers to my attention and that I support a brief discussion be included in docs. I did not mean to imply that it was my decision in any way. I am very happy to leave leading in the very capable hands of people like Allin and Jack. I hope I didn't offend anyone. Cheers Logan Logan J. Kelly, Ph.D. Associate Professor, Department of Economics, University of Wisconsin-River Falls Director, UWRF Center for Economic Research,http://www.uwrf.edu/cer Schedule a meeting: https://freebusy.io/logan.j.kelly@gmail.com Office: (715) 425-4993 Mobile: (401) 256-0986 University Email: logan.kelly(a)uwrf.edu From: Ignacio Diaz-Emparanza Sent: Friday, October 14, 8:17 AM Subject: [Gretl-devel] Hodrick-Prescott parameter To: Gretl development I see that gretl uses by default the value lambda=1600 for quarterly data, lambda = 100 for annual data and lambda=14400 for monthly data. For the quarterly case this was studied in the HP article and this value has been adopted almost as a "convention" for obtaining a trend/cycle decomposition. The values for annual and monthly data has been during some time subject to controversy. In particular, some papers consider the above values inappropiate because the agregation(from monthly to quarterly or quarterly to annual) of the trend gives very different results. But now, based on this requirement of equivalence on agregation, it seems that finally an agreement has been reached. It implies lambda(annual)=6.25 and lambda(monthly)=129600 or more generally lambda(S)=6.25*S^4 being S the seasonal periodicity. I suggest that gretl adopts these defaults. In these papers we have different justifications for such a formula: Maravall, A. y del Rı̀o, A. (2001), Time aggregation and the Hodrick- Prescott filter, Working paper 0108, Banco de España. Ravn, M. O. y Uhlig, H. (2002), ‘On adjusting the Hodrick-Prescott filter for the frequency of observations’, The Review of Economics and Statistics 84(2), 371–380. de Jong, R. M. y Sakarya, N. (2016), ‘The econometrics of the Hodrick- Prescott filter’, The Review of Economics and Statistics 98(2), 310–317. -- Ignacio Díaz-Emparanza Departamento de Economía Aplicada III (Econometría y Estadística) Universidad del País Vasco - Euskalherriko Unibertsitatea, UPV/EHU Tfno: (+34) 94 601 3732 http://www.ehu.eus/ea3 _______________________________________________ Gretl-devel mailing list Gretl-devel(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-devel _______________________________________________ Gretl-devel mailing list Gretl-devel(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-devel

To use or not to use the HP filter is something that the user must determine. In some cases,despite the drawbacks shown in these articles it may be justified. And it is a fact that many applied macroeconomists and official statisticians use it. I know that there are some literature suggesting that to obtain a good estimation of the trend one should use a data-dependent value for each variable. But there are also opinions against this, based on the necessity of comparability, for example between macroeconomic series of different regions or countries. The 'hpfilt' function we have in gretl has a fixed value for lambda, and the defaults we have do not seem to be the better. El 14/10/16 a las 15:49, Artur T. escribió: > I am not in favor of setting the default e.g. to the Ravn/Uhlig values. > Some argue that a careful approach to extract the cyclical component > would even imply a variable-depended "lambda". This doesn't rule out > adding the references Ignacio listed to the manual -- which is always a > good thing. However, if I remember correctly, Eviews offers a different > options to the user > > Just to add a note on this HP-thing: There is some critical literature > on the use of the HP-filter. Some recent ones are e.g. > > 1. Hamilton (2016): econweb.ucsd.edu/~jhamilto/hp.pdf > 2. Phillips (2015): > cowles.yale.edu/sites/default/files/files/pub/d20/d2005.pdf > > Btw, Hamilton suggests the use of a simple regression method to extract > the cyclical component for which I programmed a function package. > > Artur >

On Fri, 14 Oct 2016, Ignacio Diaz-Emparanza wrote: > The 'hpfilt' function we have in gretl has a fixed value for lambda, > and the defaults we have do not seem to be the better. Actually, you can call hpfilt with a second optional parameter: <hansl> open fedstl.bin data indpro logs indpro series hpt_l_indpro = l_indpro - hpfilt(l_indpro) series hpt_l_indpro2 = l_indpro - hpfilt(l_indpro, 129600) </hansl> ------------------------------------------------------- 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 ------------------------------------------------------- _______________________________________________ Gretl-devel mailing list Gretl-devel(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-devel

On Fri, 14 Oct 2016, Artur T. wrote:

On this I don't agree: if a choice A turns out to be more preferable than B for whatever reason, the default should be A. The only drawback in changing is, as I said, breakage in backwards compatibility.