[Gretl-users] Re: Gretl-users Digest, Vol 13, Issue 3

Re: Gretl texbook Tom asks: In my question about documentation I guess I am asking if there is a textbook that explains times series analysis as it is implemented in Gretl, i.e., is there a times series textbook where the author planned on the students using Gretl (like ITSM is used in Brockwell and Davis). One might consider Using gretl for Principles of Econometrics, 3rd edition This book is a manual for using gretl suitable for use in an introductory econometrics class. It has been written specifically for Hill, Griffiths, and Lim's Principles of Econometrics, 3rd edition, published by John Wiley and Sons, 2008. In the book, I show you how to use gretl to work all of the examples in POE. The data for the examples is packaged for installation into gretl and can be found by following the links below. (Description from the web site: http://www.learneconometrics.com/gretl.html -----Original Message----- From: gretl-users-bounces(a)lists.wfu.edu on behalf of gretl-users-request(a)lists.wfu.edu Sent: Mon 2/4/2008 3:40 AM To: gretl-users(a)lists.wfu.edu Subject: Gretl-users Digest, Vol 13, Issue 3 Send Gretl-users mailing list submissions to gretl-users(a)lists.wfu.edu To subscribe or unsubscribe via the World Wide Web, visit http://lists.wfu.edu/mailman/listinfo/gretl-users or, via email, send a message with subject or body 'help' to gretl-users-request(a)lists.wfu.edu You can reach the person managing the list at gretl-users-owner(a)lists.wfu.edu When replying, please edit your Subject line so it is more specific than "Re: Contents of Gretl-users digest..." Today's Topics: 1. Tests for Residuals (Tom La Bone) 2. Re: Tests for Residuals (Sven Schreiber) 3. Re: Re: Unit Root Test (Allin Cottrell) 4. RE: Tests for Residuals (Tom La Bone) 5. RE: Tests for Residuals (Allin Cottrell) 6. new string handling (was Re: Unit Root Test) (Allin Cottrell) 7. RE: Tests for Residuals (Allin Cottrell) 8. Re: Tests for Residuals (Sven Schreiber) ---------------------------------------------------------------------- Message: 1 Date: Sun, 3 Feb 2008 10:30:22 -0500 From: "Tom La Bone" <labone(a)gforcecable.com&gt; Subject: [Gretl-users] Tests for Residuals To: <gretl-users(a)lists.wfu.edu&gt; Message-ID: <000301c86679$b2445dc0$6401a8c0@Boozoo> Content-Type: text/plain; charset="us-ascii" Good morning, 1) Does Gretl have tests for randomness of residuals accessible as options from the GUI? 2) From the command line? 3) What references are recommended for folks who interested in the algorithms used in Gretl? In other words, what references discuss in detail how Gretl does what it does? Tom ------------------------------ Message: 2 Date: Sun, 03 Feb 2008 18:31:32 +0100 From: Sven Schreiber <svetosch(a)gmx.net&gt; Subject: Re: [Gretl-users] Tests for Residuals To: labone(a)gforcecable.com, Gretl list <gretl-users(a)lists.wfu.edu&gt; Message-ID: <47A5FA74.6080106(a)gmx.net&gt; Content-Type: text/plain; charset=ISO-8859-15; format=flowed Am 03.02.2008 16:30, Tom La Bone schrieb: Short answer: yes. Long answer: In contrast to some other open-source projects, gretl has excellent documentation; I'm almost certain the user guide has the detailed answers to your question. Again, in some cases the documentation is fairly explicit, but ultimately Allin and Jack would probably say: use the source, Luke. (Nothing I ever do since I'm not literate in C, so I'm not trying to be arrogant here, it's just the fact of open source that unlike proprietary packages everything is accessible in principle.) good luck, sven ------------------------------ Message: 3 Date: Sun, 3 Feb 2008 13:00:05 -0500 (EST) From: Allin Cottrell <cottrell(a)wfu.edu&gt; Subject: Re: [Gretl-users] Re: Unit Root Test To: Gretl list <gretl-users(a)lists.wfu.edu&gt; Message-ID: <alpine.LRH.1.00.0802031228550.309(a)ricardo.ecn.wfu.edu&gt; Content-Type: TEXT/PLAIN; charset=US-ASCII On Fri, 1 Feb 2008, Marcos Larios wrote: At present the roots are computed and printed (as you know), but they're not fetchable via a "$" variable. That probably should be added. In the meantime this might be helpful: if you estimate an ar(i)ma model using the --x-12-arima option in gretl, a plain text file named varname.rts (where "varname" is the name of the dependent variable) is written. This is in a simple fixed format and contains all the roots information. It will be written to your x12arima working directory. You can find out where that is using print "@x12adir" in gretl. We currently have code for graphing the roots when you estimate a VAR. You can see how we do this in either of these ways. (a) Estimate a arbitrary VAR in the GUI. In the VAR output window select Graphs/VAR inverse roots. Click on the graph and select "Save to session as icon". Open the icon view window (View/Icon view). Right-click on the graph icon and select "Edit plot commands". You'll get a window showing the gnuplot commands for making this sort of graph, which you could save under another name and use as a template. (b) Look at the C code that constructs the graph commands for this case. http://gretl.cvs.sourceforge.net/*checkout*/gretl/gretl/lib/src/graphing.... and look for "gretl_VAR_roots_plot". Allin Cottrell ------------------------------ Message: 4 Date: Sun, 3 Feb 2008 13:04:42 -0500 From: "Tom La Bone" <labone(a)gforcecable.com&gt; Subject: RE: [Gretl-users] Tests for Residuals To: "'Sven Schreiber'" <svetosch(a)gmx.net&gt;, "'Gretl list'" <gretl-users(a)lists.wfu.edu&gt; Message-ID: <000401c8668f$41aaacc0$6401a8c0@Boozoo> Content-Type: text/plain; charset="us-ascii" Sven, thanks for your timely response. In case I have not blown my cover already, I must confess that I am a stat grad student taking his first time series course. Our professor has given us the freedom to use any software we want for the course, and I am taking the opportunity to learn Gretl. I chose Gretl because of many of the things you mention, i.e., it is well documented, open source (I can look at the code if I dare), and in the worst case it is a nice front-end for sending data to R for further analysis. The main downside of using Gretl is that nobody in the department is familiar with the program so I am "on my own" in my efforts to learn how to use it. I make an honest effort to look through the manual for answers to my questions before asking for help. I have gotten some useful answers and guidance that were not immediately obvious from the manual. On the other hand I have gotten some guidance that goes into the "I have not gotten to that chapter yet" bin. It would be helpful to be able to search the archives for answers to my questions that have already been posted, but it appears that is not possible. In my question about documentation I guess I am asking if there is a textbook that explains times series analysis as it is implemented in Gretl, i.e., is there a times series textbook where the author planned on the students using Gretl (like ITSM is used in Brockwell and Davis). Tom -----Original Message----- From: Sven Schreiber [mailto:svetosch@gmx.net] Sent: Sunday, February 03, 2008 12:32 PM To: labone(a)gforcecable.com; Gretl list Subject: Re: [Gretl-users] Tests for Residuals Am 03.02.2008 16:30, Tom La Bone schrieb: Short answer: yes. Long answer: In contrast to some other open-source projects, gretl has excellent documentation; I'm almost certain the user guide has the detailed answers to your question. detail Again, in some cases the documentation is fairly explicit, but ultimately Allin and Jack would probably say: use the source, Luke. (Nothing I ever do since I'm not literate in C, so I'm not trying to be arrogant here, it's just the fact of open source that unlike proprietary packages everything is accessible in principle.) good luck, sven ------------------------------ Message: 5 Date: Sun, 3 Feb 2008 16:47:34 -0500 (EST) From: Allin Cottrell <cottrell(a)wfu.edu&gt; Subject: RE: [Gretl-users] Tests for Residuals To: labone(a)gforcecable.com, Gretl list <gretl-users(a)lists.wfu.edu&gt; Message-ID: <alpine.LRH.1.00.0802031640270.309(a)ricardo.ecn.wfu.edu&gt; Content-Type: TEXT/PLAIN; charset=US-ASCII On Sun, 3 Feb 2008, Tom La Bone wrote: OK, then we'll try to be extra-helpful ;-) 1) Does Gretl have tests for randomness of residuals accessible as options from the GUI? Yes, various aspects of the "randomness" (or otherwise) of the residuals are explored by almost all the items under the "Tests" menu in the model output window. You could also save the residuals and do a runs test if you wanted. 2) From the command line? help lmtest help testuhat help runs 3) What references are recommended for folks who interested in the algorithms used in Gretl? In other words, what references discuss in detail how Gretl does what it does? The Gretl User's Guide, latest version at http://ricardo.ecn.wfu.edu/pub/gretl/manual/ plus of course the C source, as Sven said. Allin Cottrell ------------------------------ Message: 6 Date: Sun, 3 Feb 2008 21:15:41 -0500 (EST) From: Allin Cottrell <cottrell(a)wfu.edu&gt; Subject: [Gretl-users] new string handling (was Re: Unit Root Test) To: Gretl list <gretl-users(a)lists.wfu.edu&gt; Message-ID: <alpine.LRH.1.00.0802032105470.3269(a)ricardo.ecn.wfu.edu&gt; Content-Type: TEXT/PLAIN; charset=US-ASCII On Sun, 3 Feb 2008, I wrote: Around Christmas I added some functionality for handling strings. I think the timing was bad and it wasn't much noticed, and as a result was not much debugged. However, reading roots information from a X-12-ARIMA .rts file is a handy case in point. I've had a go at this, and in the process have fixed a few bugs and added some more functionality. Find below a function that reads an .rts file and returns a roots matrix. The new string-handling idioms are explained in (a) "help sscanf" and (b) section 11.2 of the User's Guide (CVS and Windows snapshot). <script> function read_x12a_roots (series y, int type[1:2:1]) # compose the name of the X12A roots file... sprintf yfile, "%s/%s.rts", @x12adir, argname(y) # and grab its content into a string string rts = readfile(@yfile) matrix r matrix R if (type = 1) string targ = "AR" else string targ = "MA" endif string line = strstr(@rts, @targ) loop while isstring(line) --quiet # offset on line to find the numbers string line = @line + 18 # get 4 values: real, imaginary, modulus, frequency sscanf @line, "%m", r R = R|r string line = strstr(@line, @targ) endloop return matrix R end function open data9-7 arma 2 1 ; QNC --x-12-arima AR_roots = read_x12a_roots(QNC, 1) MA_roots = read_x12a_roots(QNC, 2) print AR_roots MA_roots </script> Allin. ------------------------------ Message: 7 Date: Sun, 3 Feb 2008 21:51:24 -0500 (EST) From: Allin Cottrell <cottrell(a)wfu.edu&gt; Subject: RE: [Gretl-users] Tests for Residuals To: labone(a)gforcecable.com, Gretl list <gretl-users(a)lists.wfu.edu&gt; Message-ID: <alpine.LRH.1.00.0802032147370.3269(a)ricardo.ecn.wfu.edu&gt; Content-Type: TEXT/PLAIN; charset=US-ASCII On Sun, 3 Feb 2008, Tom La Bone wrote: Try Lee Adkins's site: http://www.learneconometrics.com/gretl.html Allin Cottrell ------------------------------ Message: 8 Date: Mon, 04 Feb 2008 09:39:48 +0100 From: Sven Schreiber <svetosch(a)gmx.net&gt; Subject: Re: [Gretl-users] Tests for Residuals To: Gretl list <gretl-users(a)lists.wfu.edu&gt; Message-ID: <47A6CF54.5080202(a)gmx.net&gt; Content-Type: text/plain; charset=ISO-8859-15; format=flowed Am 04.02.2008 03:51, Allin Cottrell schrieb: Very interesting link; maybe worth adding it (or even an entire section with related stuff) to the website, if it's not there already? -sven ------------------------------ _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users End of Gretl-users Digest, Vol 13, Issue 3 ******************************************