2009/10/2 Allin Cottrell <cottrell(a)wfu.edu&gt; I'm a little bit confused with the line "cm" (line 10292 in pt_BR.po). Is it about "centimeter"? Any comments, not-yet-reported bugs that people want to see fixed I think it would be nice if we could improve Gretl (for Mac) icon size. I think Tiger and previous OS' the maximum icon size was 128x128 but in Leopard they go up to 512X512. But of course this is not a big deal, just cosmetics ;) Another comment: when we open the Gretl.dmg file (to install Gretl on Mac) a README.pdf outdated file is shown (it refers to a 1.7.5 version). Best regards, Henrique C. de Andrade Doutorando em Economia Aplicada Universidade Federal do Rio Grande do Sul www.ufrgs.br/ppge

Allin Cottrell schrieb: I'm using this opportunity to point out the feature requests on the tracker that are more than one year old. Here's a list with some thoughts: (I just closed the VECM-related feature request, see my comment there.) * Improvement of logit regression modelling in gretl - ID: 1886339 E.g. multinomial logit -- this is of course a big thing and should just wait I guess. * access to the ranking() command from the GUI - ID: 1906097 IMHO there is no reason not to decide on what to do with it now, no? Is there a plan to implement it (probably in the vicinity of 'add logs' etc.) or should it be discarded? * Saving rankings from difftest () to the data set - ID: 1906103 Not sure what the use case is. Would it be enough to implement the previous feature? Actually if the previous feature is implemented I guess it would be relatively easy to write a function package that covers this request. So in that sense I would tend to close it, adding an appropriate comment suggesting such a package. BTW, in the test output I found some strings that are not yet marked for translation: string: (zero differences: 0, non-zero ties: 0) string: Two-tailed p-value = 0.2 strings: value rank group * gnuplot --loess-fit - ID: 2000792 more precisely, this is about getting date labels on the x-axis in combination with --loess-fit. Here I would also tend to outsource it to someone with sufficient gnuplot knowledge who could make a function package for it. Or is this 'solution' too lazy and it really should be done in gretl proper? * $hausman should also access random vs. fixed effects test - ID: 2034862 well this is a request by myself, so of course I think it should be implemented :-) cheers, sven

On Fri, 2 Oct 2009, Sven Schreiber wrote: As Jack said, we now do multinomial logit. That has been available since gretl 1.8.1 (May, 2009). I'm not sure it's worth implementing the remaining points: if the user wants a reference category other than the default (i.e. the lowest value), it's easy enough to redefine the dependent variable accordingly, as in alt_y = (y == 5)? 0 : y to make (y = 5) the reference category (assuming there isn't already a value 0 in the y series). I'm not very enthusiastic. If we add rankings to the GUI menu we'd presumably have to add sorted series also, and my feeling is that this would clutter the menu. I like the idea of hiving this off to a package. OK, I'll take a look at that. The thing is that we don't offer the --loess-fit option (or in fact, any fit option) for time-series plots. You can get such a plot by doing genr time plot y time --loess But then it's quite right that the x-axis values are 1, 2, 3, ... because those are the values of the variable called "time". So I suppose this one is really a request to implement the gnuplot "fit" options for time-series plots. Yes, that one should be implemented -- and now it is, in CVS. Allin.

On Fri, 2 Oct 2009, Allin Cottrell wrote: Last week I taught a summer school where all the students were asked to use gretl; all of them but two were using windows. I was absolutely astonished by the fact that, from time to time, the gui client would crash and they'd just go "oh, well", and start their work again. I'd ask "what were you doing when gretl crashed?" and, in most cases, only got vague answers. (Footnote: a very annoying feature of point-n-click software is that some bugs are _way_ more difficult to reproduce and hence to fix). Of course gretl should never crash, but Allin has always done everything humanly possible to prevent this (although the fact that development takes place on linux doesn't help) doing his stellar job as per usual. The essential ingredient we lack in these cases is bug reports. I understand that the average windows user doesn't even consider the idea of filing a bug report. This happens for a number of reasons I won't delve into, but please please pretty please people, if gretl ever crashes on you, you can do yourselves, the developers and the community at large a big favour: 1) Spend some time to isolate the exact action that makes gretl crash. That is, try to make gretl crash again by re-doing your actions. 2a) If you find a way to trigger a crash, that's excellent. Tell it to the list. It usually takes AC days if not hours to fix these things. 2b) If you _don't_ find a way to reproduce the crash consistently, well tell us all the same! User-interface crashes can be very elusive, and may depend on a number of things, but just delimiting the search (like, for example "I was writing a script" or "I was estimating an ARIMA model") may help a lot. Remember guys, when you use gretl you get a good econometrics package for free. Of course, no-one _expects_ you to pay anything back, but just filing a bug report is an excellent way to contribute to the project, much more important than most people realise. Riccardo (Jack) Lucchetti Dipartimento di Economia Università Politecnica delle Marche r.lucchetti(a)univpm.it http://www.econ.univpm.it/lucchetti

I have a few bug reports and suggestions before the 1.8.5 myself: 1)- Gretl crashes choosing the "XY Scatter Plot" option from the right click menu. 2)- The little "^" signs for sorting the variables based on ID# etc seem a little strange. I don't know if it is just me but I see three "^" signs for the three columns visable at all times. I would normally expect to see just one of them, since there is only one sort criteria. 3) I agree that there are enough bug fixes and feature improvements for a new release but I feel that some of them still need some finishing touches. I think it would be really nice to relase gretl 1.8.5 after perfecting the new GUI features with a few minor additions such as: a)- Adding a "restore model sample" option in the Edit menu for the model window so that the user does not have to close and reopen model windows just to get the newly implemented and convenient dialog box for restoring the sample option. b)- I still don't see why saving fitted values, residuals and squared residuals need to be disabled while the various tests can be very neatly done without reverting to the model sample. c)- I think it would be complementary to the refinements for the GUI "model table" to improve the LaTeX output of the "summary statisitcs" table as well. In particular, it would be very nice to be able to set the number of significant digits and also disable some statistics in favor of a one-part table. 4)- I second Sven's request to implement the gnuplot "fit" options for time-series plots. In fact here is a suggestion: Since we now have a separate "lines" tab, how about removing fitted line option from the main tab and instead add a small icon next to the formula line for specifying a new line? This new icon should give a drop down list of the various available fitted line options including a few more such a heteroscedasticity corrected ols fitted line. This way we can easily create graphs with more than one fitted line. I think this would be powerful. I hope these are considered before the new release. Cheers A. Talha Yalta On Fri, Oct 2, 2009 at 5:29 PM, Allin Cottrell <cottrell(a)wfu.edu&gt; wrote: -- “Remember not only to say the right thing in the right place, but far more difficult still, to leave unsaid the wrong thing at the tempting moment.” - Benjamin Franklin (1706-1790) --

This is strange. I cannot be more specific since there is nothing special. I open a data set and select 2 variables. Selecting xy scatter crashes gretl after trying to set the x axis variable. Other plots seem normal. Running gretl from the console does not give any messages before crashing either. I was running my freshly baked cvs compilation from this morning. -- “Remember not only to say the right thing in the right place, but far more difficult still, to leave unsaid the wrong thing at the tempting moment.” - Benjamin Franklin (1706-1790) --

On Mon, Oct 5, 2009 at 1:16 PM, Sven Schreiber <svetosch(a)gmx.net&gt; wrote: I fully agree. Talha -- “Remember not only to say the right thing in the right place, but far more difficult still, to leave unsaid the wrong thing at the tempting moment.” - Benjamin Franklin (1706-1790) --

On Mon, 5 Oct 2009, Talha Yalta wrote: As I mentioned, this is now fixed. What version of GTK do you have on your system? I see only one "sort arrow" at any given time, depending on the column selected, and no arrows if you return the sorting to its "base state" (arranged by ID number). Possibly. That's because you haven't looked at the code ;-). Or in other words, you haven't had to think through the full implications of adding new variables in the case where we're looking at two different sub-sampled datasets: the one the model was estimated on, and the one currently in force. (It's not so difficult if the full dataset is currently in force and only the model's dataset was sub-sampled, but even that is tricky to get right.) No doubt, but I'm not going to make a 1.8.5 release conditional on that. If you produce a single time-series plot in the GUI you now get the option of adding a linear, quadratic or loess fit against time, with dates appearing on the x-axis. But this needs more testing. Maybe later -- though it sounds a bit over-complicated to me -- but not for 1.8.5. Allin

On Mon, Oct 5, 2009 at 5:59 PM, Allin Cottrell <cottrell(a)wfu.edu&gt; wrote: Checking my system, I see that I have gtk2 2.16.6 installed in my Pardus 2009 system. The window manager is KDE 4.2. I have 2 more issues to report: 1)- When I add a new line from a formula and choose its appearance as a colorless dashed line, adding a second line from a formula destroys the color selector of the first line and makes it permanently dashed. 2)- On my Linux box I keep getting comma as the decimal separator both in normal and LaTeX outputs. I cannot get dots whether or not gretl is set to English. When the language is set to Turkish, starting gretl from the console with the "gretl --english" command gives in the console setlocale: 'tr_TR' -> 'tr_TR' Also, whenever I choose English from the preferences, there is the console output setlocale: 'C' -> 'C' I don't know if the above output is normal or gives any hint on the source of the problem. I was able to check the latest snapshot installed on a Turkish Windows XP system and there are no problems there. Do you have a suggestion? Cheers Talha -- “Remember not only to say the right thing in the right place, but far more difficult still, to leave unsaid the wrong thing at the tempting moment.” - Benjamin Franklin (1706-1790) --

On Tue, 6 Oct 2009, Talha Yalta wrote: That's now fixed in CVS and snapshot. Allin.

On Tue, 6 Oct 2009, Henrique wrote: I think my last upload must not have completed successfully, so you only got part of the file. I'll try again this evening. Allin

On Mon, 5 Oct 2009, Henrique wrote: I've just tried the current snapshot on Vista. There, the sort arrows are shown facing horizontally to indicate that a column is not currently the sort criterion, otherwise they face up or down. They never disappear altogether. This is not the behavior I see on Linux. It is part of the GTK Windows-emulation theme; that is, it's supposed to resemble the behavior of sort arrows in native Windows apps. You can turn off Windows emulation under Preferences if you wish. Or you could try researching relevant settings in the gtkrc file, which lives in <gretldir>/share/themes/MS-Windows/gtk-2.0/ . I don't know if you can change the sort-arrow appearance at that level, but maybe you can. Allin Cottrell

On Mon, 5 Oct 2009, Ignacio Diaz-Emparanza wrote: No, it's actually "pane", as in "window pane". In gretl CVS, you can split the output window in two, to make it easier to see different parts of the output script (possibly far away from each other) simultaneously. This is a way of working that we aristocrats (ie Emacs users) are quite fond of, although I personally like it even better when the split is vertical. Riccardo (Jack) Lucchetti Dipartimento di Economia Università Politecnica delle Marche r.lucchetti(a)univpm.it http://www.econ.univpm.it/lucchetti

Allin Cottrell schrieb: I have just committed the updated de.po. At least gretl doesn't crash with it right away, so I hope it's ok. good luck for the release! cheers, sven