From ???@??? Wed Jan 19 08:48:18 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id GAA24428 for ; Wed, 19 Jan 2000 06:50:04 +1100 Message-Id: <200001181950.GAA24428@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Wed, 19 Jan 2000 06:50:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 19 Jan 100 06:46:54 GMT+1000 From: James Peery Cover To: "RATS Discussion List" Subject: GARCH and regime switch Date: Tue, 18 Jan 2000 13:47:42 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: University of Alabama X-Mailer: Mozilla 4.04 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear Fellow RATS users, Does anyone know whether it is possible to test for a change in regime if one is assuming a GARCH process for the error term? I have a two equation model with two dependent variables each following AR(1) processes with GARCH errors. I suspect a change in regime at a certain point in time. Is there a formal test that can be applied? -- James Peery Cover Department of Economics and Finance Culverhouse College of Commerce University of Alabama P. O. Box 870224 Tuscaloosa, AL 35487-0224 From ???@??? Thu Jan 20 08:43:36 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id FAA14316 for ; Thu, 20 Jan 2000 05:00:05 +1100 Message-Id: <200001191800.FAA14316@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Thu, 20 Jan 2000 05:00:05 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 20 Jan 100 04:40:08 GMT+1000 From: Jim Lee To: "RATS Discussion List" Subject: Re: GARCH and regime switch Date: Wed, 19 Jan 2000 11:40:52 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: TAMU-CC X-Mailer: Mozilla 4.61 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit James: A test for this purpose has been developed by James Chu in: "Detecting Parameter Shift in GARCH Models," Econometric Reviews, 14 (2), 1995, 241-266. James Peery Cover wrote: > Dear Fellow RATS users, > Does anyone know whether it is possible to test for a change in regime > if one is assuming a GARCH process for the error term? I have a two > equation model with two dependent variables each following AR(1) > processes with GARCH errors. I suspect a change in regime at a certain > point in time. Is there a formal test that can be applied? > > -- > James Peery Cover > Department of Economics and Finance > Culverhouse College of Commerce > University of Alabama > P. O. Box 870224 > Tuscaloosa, AL 35487-0224 -- Jim Lee Texas A&M University-Corpus Christi From ???@??? Thu Jan 20 12:32:34 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id MAA08396 for ; Thu, 20 Jan 2000 12:15:04 +1100 Message-Id: <200001200115.MAA08396@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Thu, 20 Jan 2000 12:15:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 20 Jan 100 12:13:33 GMT+1000 From: Feng Zhu To: "RATS Discussion List" Subject: simulations and frequency domain analysis Date: Wed, 19 Jan 2000 20:13:37 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.40 Dear everyone, I am trying to run the RATS program and have some major difficulties. I would be very grateful if you could help me with simulations and running regressions in the frequency domain., 1. How to generate a large no. of random no.s and save these for later use. How to establish replications of the regression? 2. How to design a loop to run simulations and save the results for analysis. 3. How to design a procedure to compute a sum such as: som from k= 1 to m, of the forward factorial (d)k=d*(d+1)*...*(d+k-1), multiplid by u(t-k), random no., and divided by k!. It might be easier ro use a gama function in the sum? 4. The definition in FFT is different from mine, how can I modify an instruction? How do I run, say, a log-periodogram regression with RATS? Thanks for your help! Zhu Feng From ???@??? Thu Jan 20 13:12:44 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id MAA08150 for ; Thu, 20 Jan 2000 12:50:04 +1100 Message-Id: <200001200150.MAA08150@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Thu, 20 Jan 2000 12:50:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 20 Jan 100 12:46:40 GMT+1000 X-UIDL: 948333674.012 From: Feng Zhu To: "RATS Discussion List" Subject: simulations and frequency domain analysis Date: Wed, 19 Jan 2000 20:46:16 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.40 Status: U Dear everyone, I am trying to run the RATS program and have some major difficulties. I would be very grateful if you could help me with simulations and running regressions in the frequency domain., 1. How to generate a large no. of random no.s and save these for later use. How to establish replications of the regression? 2. How to design a loop to run simulations and save the results for analysis. 3. How to design a procedure to compute a sum such as: som from k= 1 to m, of the forward factorial (d)k=d*(d+1)*...*(d+k-1), multiplid by u(t-k), random no., and divided by k!. It might be easier ro use a gama function in the sum? 4. The definition in FFT is different from mine, how can I modify an instruction? How do I run, say, a log-periodogram regression with RATS? Thanks for your help! Zhu Feng From ???@??? Fri Jan 21 09:39:31 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id RAA22510 for ; Thu, 20 Jan 2000 17:00:05 +1100 Message-Id: <200001200600.RAA22510@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Thu, 20 Jan 2000 17:00:05 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 20 Jan 100 16:58:57 GMT+1000 From: Petri Rouvinen To: "RATS Discussion List" Subject: unsubscribe Date: Thu, 20 Jan 2000 08:00:23 +0200 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32) (via Mercury MTS (Bindery) v1.40) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" unsubscribe From ???@??? Fri Jan 21 09:40:46 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id BAA15556 for ; Fri, 21 Jan 2000 01:00:04 +1100 Message-Id: <200001201400.BAA15556@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 21 Jan 2000 01:00:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 21 Jan 100 00:41:09 GMT+1000 From: "Christopher F. Baum" To: "RATS Discussion List" Subject: Re: simulations and frequency domain analysis Date: Thu, 20 Jan 2000 08:41:10 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: X-Mailer: Mulberry (MacOS) [1.4.5, s/n P020-300786-009] (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit An illustration of how to run a log-periodogram regression is in Estima's GPH.SRC. An extension to handle seasonally adjusted data, GPH_SEAS, is available at http://ideas.uqam.ca/ideas/data/Softwares/bocbocodeR980223.html in the SSC-IDEAS archive. All RATS-L readers are invited to contribute RATS code to be shared to the SSC-IDEAS archive. Kit Baum SSC-IDEAS archive maintainer http://ideas.uqam.ca/ideas/data/bocbocode.html --On Wed, Jan 19, 2000 20:46 -0500 Feng Zhu wrote: > Dear everyone, > > I am trying to run the RATS program and have some major difficulties. I > would be very grateful if you could help me with simulations and running > regressions in the frequency domain., > > 1. How to generate a large no. of random no.s and save these for later > use. How to establish replications of the regression? > 2. How to design a loop to run simulations and save the results for > analysis. > 3. How to design a procedure to compute a sum such as: > som from k= 1 to m, of the forward factorial (d)k=d*(d+1)*...*(d+k-1), > multiplid by u(t-k), random no., and divided by k!. It might be easier > ro use a gama function in the sum? > 4. The definition in FFT is different from mine, how can I modify an > instruction? How do I run, say, a log-periodogram regression with RATS? > > Thanks for your help! > > Zhu Feng > > > ----------------------------------------------------------------- Kit Baum baum@bc.edu http://fmwww.bc.edu/ec-v/baum.fac.html From ???@??? Fri Jan 21 09:42:13 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id EAA12290 for ; Fri, 21 Jan 2000 04:00:05 +1100 Message-Id: <200001201700.EAA12290@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 21 Jan 2000 04:00:05 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 21 Jan 100 03:51:55 GMT+1000 From: "Estima" To: "RATS Discussion List" Subject: Re: simulations and frequency domain analysis Date: Thu, 20 Jan 2000 10:42:59 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Estima MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v3.11) (via Mercury MTS (Bindery) v1.40) > > 1. How to generate a large no. of random no.s and save these for later > use. Have you read Chapter 10 of the manual? If not, take a look--much of this information is provided there. > How to establish replications of the regression? > 2. How to design a loop to run simulations and save the results for > analysis. Again, Chapter 10 provides details on the bookkeeping required. See the RATS FAQ on our web page for an example of doing rolling regressions. Although you are talking about doing repetitions of the same estimation, rather than moving through time periods, many of the basic concepts apply. > 3. How to design a procedure to compute a sum such as: > som from k= 1 to m, of the forward factorial (d)k=d*(d+1)*...*(d+k-1), > multiplid by u(t-k), random no., and divided by k!. It might be easier > ro use a gama function in the sum? There's no built in factorial function in RATS (I believe we've added one in the upcoming Version 5). To get N!, use the expression: exp(%lngamma(N+1)) > 4. The definition in FFT is different from mine, how can I modify an > instruction? How do I run, say, a log-periodogram regression with RATS? You can't really "modify" a RATS instruction per se, beyond the flexibility provided by the various parameters and options. However, the spectral instructions are designed to be fairly "low level" tools that provide a great deal of flexibility. As Kit Baum suggested, see GPH.SRC or his GPH_SEAS for some examples. Sincerely, Tom Maycock -- ------------------------------------------------------------ | Estima | Sales: (800) 822-8038 | | P.O. Box 1818 | (847) 864-8772 | | Evanston, IL 60204-1818 | Support: (847) 864-1910 | | USA | Fax: (847) 864-6221 | | http://www.estima.com | estima@estima.com | ------------------------------------------------------------ From ???@??? Fri Jan 21 10:04:20 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id KAA02388 for ; Fri, 21 Jan 2000 10:00:04 +1100 Message-Id: <200001202300.KAA02388@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 21 Jan 2000 10:00:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 21 Jan 100 09:58:50 GMT+1000 From: Feng Zhu To: "RATS Discussion List" Subject: Re: simulations and frequency domain analysis Date: Thu, 20 Jan 2000 17:58:36 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.40 Thanks for your help. I am still trying to figure out the right way to do the work. I will let you know if I have furthe difficulties! All the best! Zhu On Thu, 20 Jan 2000, Christopher F. Baum wrote: > An illustration of how to run a log-periodogram regression is in Estima's > GPH.SRC. An extension to handle seasonally adjusted data, GPH_SEAS, is > available at > > http://ideas.uqam.ca/ideas/data/Softwares/bocbocodeR980223.html > > in the SSC-IDEAS archive. > > All RATS-L readers are invited to contribute RATS code to be shared to the > SSC-IDEAS archive. > > Kit Baum > SSC-IDEAS archive maintainer > http://ideas.uqam.ca/ideas/data/bocbocode.html > > --On Wed, Jan 19, 2000 20:46 -0500 Feng Zhu wrote: > > > Dear everyone, > > > > I am trying to run the RATS program and have some major difficulties. I > > would be very grateful if you could help me with simulations and running > > regressions in the frequency domain., > > > > 1. How to generate a large no. of random no.s and save these for later > > use. How to establish replications of the regression? > > 2. How to design a loop to run simulations and save the results for > > analysis. > > 3. How to design a procedure to compute a sum such as: > > som from k= 1 to m, of the forward factorial (d)k=d*(d+1)*...*(d+k-1), > > multiplid by u(t-k), random no., and divided by k!. It might be easier > > ro use a gama function in the sum? > > 4. The definition in FFT is different from mine, how can I modify an > > instruction? How do I run, say, a log-periodogram regression with RATS? > > > > Thanks for your help! > > > > Zhu Feng > > > > > > > > > > ----------------------------------------------------------------- > Kit Baum baum@bc.edu http://fmwww.bc.edu/ec-v/baum.fac.html > From ???@??? Sat Jan 22 09:28:06 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id CAA22396 for ; Sat, 22 Jan 2000 02:00:04 +1100 Message-Id: <200001211500.CAA22396@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 22 Jan 2000 02:00:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 22 Jan 100 01:37:00 GMT+1000 From: Simon.Van-Norden@hec.ca To: "RATS Discussion List" Subject: Roots of AR Polynomial Date: Fri, 21 Jan 2000 09:40:16 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Ecole des Hautes Etudes Commerciales X-Mailer: Mozilla 4.61 [en] (Win98; U) (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit I've got an AR(3) model and I need to find the roots of the AR polynominal. Can anyone tell me how to do it in RATS? Just to be clear, I have (1 - rho1*L - rho2*L^2 - rho3*L^3) * y(t) = e(t) and I need to factor it into the form (1-phi1*L)*(1-phi2*L)*(1-phi3*L)*y(t) = e(t) How do I find the phi's given the rho's? I'd be grateful if someone can point me towards the right function or matrix formula. Regards, SvN -- Simon van Norden, Prof. agrŽgŽ, www.hec.ca/pages/simon.van-norden Service de l'enseignement de la finance, ƒcole des H.E.C. 3000 Cote-Sainte-Catherine, Montreal QC, CANADA H3T 2A7 simon.van-norden@hec.ca or (514)340-6781 or fax:(514)340-5632 From ???@??? Sat Jan 22 09:28:20 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id CAA12976 for ; Sat, 22 Jan 2000 02:30:04 +1100 Message-Id: <200001211530.CAA12976@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 22 Jan 2000 02:30:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 22 Jan 100 02:27:37 GMT+1000 From: Sune Karlsson To: "RATS Discussion List" Subject: Re: Roots of AR Polynomial Date: Fri, 21 Jan 2000 16:28:52 +0100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: MIME-version: 1.0 X-Mailer: Mozilla 4.61 [en] (WinNT; I) (via Mercury MTS (Bindery) v1.40) Content-type: text/plain Content-Transfer-Encoding: 8bit Simon.Van-Norden@hec.ca wrote: > > I've got an AR(3) model and I need to find the roots of the AR polynominal. Can > anyone tell me how to do it in RATS? > > Just to be clear, I have > (1 - rho1*L - rho2*L^2 - rho3*L^3) * y(t) = e(t) > and I need to factor it into the form > (1-phi1*L)*(1-phi2*L)*(1-phi3*L)*y(t) = e(t) > How do I find the phi's given the rho's? > > I'd be grateful if someone can point me towards the right function or matrix > formula. One way to do this is to write the process as a vector AR(1) process x(t) = A*x(t) + u(t) with x(t) = {y(t),y(t-1),y(t-2)}', rho1 rho2 rho3 A = 1 0 0 0 1 0 and solve for the eigenvalues of A, this will give you the phis. /Sune > Regards, > > SvN > -- > Simon van Norden, Prof. agrŽgŽ, www.hec.ca/pages/simon.van-norden > Service de l'enseignement de la finance, ƒcole des H.E.C. > 3000 Cote-Sainte-Catherine, Montreal QC, CANADA H3T 2A7 > simon.van-norden@hec.ca or (514)340-6781 or fax:(514)340-5632 -- Sune Karlsson | Fax: + 46 8 34 81 61 Stockholm School of Economics | Phone: + 46 8 736 92 39 Box 6501, 113 83 Stockholm, Sweden | http://www.hhs.se/personal/SuneK/ http://www.hhs.se/stat/ | http://swopec.hhs.se/ From ???@??? Sat Jan 22 09:28:44 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id FAA15456 for ; Sat, 22 Jan 2000 05:30:04 +1100 Message-Id: <200001211830.FAA15456@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 22 Jan 2000 05:30:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 22 Jan 100 05:16:54 GMT+1000 From: "Norman J Morin" To: "RATS Discussion List" Subject: Re: Roots of AR Polynomial Date: Fri, 21 Jan 2000 13:15:33 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-type: text/plain; charset=us-ascii X-Mailer: Mercury MTS (Bindery) v1.40 Hello, I wrote this a while back, and don't remember it well, but I think it does the trick. Norm ******************************************************************************** PROCEDURE FACTOR invec outvec ******************************************************************************** * * Factors 1 - a(1)*L - a(2)*L^2 - ... - a(p)*L^p * * into (1-b(1)*L)*(1-b(2)*L)*...*(1-b(p)*L) * * where [a(1),...,a(p)] = invec * and outvec = [b(1),...,b(p)] * * [PRINT]/NOPRINT Print output or not * BDROP = [0] Number of entires to drop at the beginning of the * input vector * EDROP = [0] Number of entires to drop at the end of the * input vector * * The bdrop and edrop command allows one to use the %beta vector, which is * output from the LINREG command, as the input vector when there * are other variables included with lagged dependent variables. * * E.g., for * LINREG y * #constant y{1 to 4} dum1974 dum1982 * one would use * @FACTOR(bdrop=1,edrop=2) %beta roots * * By Norman Morin (m1njm00@frb.gov), 7/97 * * ******************************************************************************** OPTION SWITCH PRINT 1 OPTION INTEGER BDROP 0 OPTION INTEGER EDROP 0 TYPE VECTOR invec *outvec LOCAL INTEGER p LOCAL VECT newvec LOCAL RECT F DIM newvec(%ROWS(invec)-edrop-bdrop) EWISE newvec(i) = invec(i+bdrop) COMPUTE p = %ROWS(newvec) DIM F(p,p) COMPUTE F = %CONST(0.0) DO i = 1,p DO j = 1,p IF j.eq.(i-1) COMPUTE F(i,j) = 1. END DO COMPUTE F(1,i) = newvec(i) IF i.ne.1 COMPUTE F(i,p) = 0. END DO EIGEN(cvalues=cvec) F outvec IF PRINT.eq.1 { DISP DISP 'Factoring the polynomial:' DISP ######## 1 DO jj = 1,p DISP ' -' ###.##### newvec(jj) "* L^" @-1 jj END DISP DISP ' The roots of the polynomial in the lag operator: ' DO jj = 1,p IF abs(%imag(cvec(jj))).lt.(.00000001) DISP ########.##### %real(cvec(jj)) ELSE IF abs(%real(cvec(jj))).lt.(.00000001) DISP %imag(cvec(jj)) @-1 "*i" ELSE DISP ########.##### %real(cvec(jj)) @-1 ' + ' @-1 $ ##.##### %imag(cvec(jj)) @-1 "*i" @33 '(' $ @-1 #.##### (%real(cvec(jj))**2+%imag(cvec(jj))**2)**.5 @-1 ')' END DO DISP DISP ' implying the roots of the difference equation:' DO jj = 1,p IF abs(%imag(1./cvec(jj))).lt.(.00000001) DISP ########.##### %real(1./cvec(jj)) ELSE IF abs(%real(1./cvec(jj))).lt.(.00000001) DISP %imag(1./cvec(jj)) @-1 "*i" ELSE DISP ########.##### %real(1./cvec(jj)) @-1 ' + ' @-1 $ ##.##### %imag(1./cvec(jj)) @-1 "*i" @33 '(' $ @-1 #.##### (%real(1./cvec(jj))**2+%imag(1./cvec(jj))**2)**.5 @-1 ')' END DO DISP } END -- Norman J. Morin --------------------------------------------------------- Division of Research and Statistics * Mail Stop 82 Board of Governors of the Federal Reserve System Washington, D.C. 20551 * (202) 452-2476 norman.j.morin@frb.gov * nmorin@frb.gov * m1njm00@frb.gov --------------------------------------------------------- From ???@??? Tue Jan 25 08:42:58 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id DAA19720 for ; Tue, 25 Jan 2000 03:00:05 +1100 Message-Id: <200001241600.DAA19720@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Tue, 25 Jan 2000 03:00:05 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 25 Jan 100 02:49:08 GMT+1000 From: "Karen C. Seto" To: "RATS Discussion List" Subject: unbalanced panels Date: Mon, 24 Jan 2000 10:49:24 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.40 Does anyone know if and how to estimate unbalanced panels in rats? Many thanks, Karen Seto -- Karen C. Seto Boston University Center for Energy and Environmental Studies, Department of Geography 675 Commonwealth Avenue, Room 141E Boston, MA 02215 617-353-7554 (voice) 617-353-5986 (fax) From ???@??? Thu Jan 27 09:52:51 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id GAA11864 for ; Thu, 27 Jan 2000 06:40:06 +1100 Message-Id: <200001261940.GAA11864@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Thu, 27 Jan 2000 06:40:06 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 27 Jan 100 06:39:18 GMT+1000 From: Simon.Van-Norden@hec.ca To: "RATS Discussion List" Subject: Postscript Output of Graphs Date: Wed, 26 Jan 2000 14:42:41 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Ecole des Hautes Etudes Commerciales X-Mailer: Mozilla 4.61 [en] (Win98; U) (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit I'm using RATS graphics for the first time in a long time and I'm having trouble getting properly sized Postscript output. I want to place a stacked graph I created with SPGRAPH on a page in Portrait orientation. When I print direct from GRAFEDIT, the bottom third of the page is left blank. When I export from GRAFEDIT to postscript (Portrait), I get the top third of the page left blank. Is there any way to get RATS to proportion the page more realistically? Thanks, SvN -- Simon van Norden, Prof. agrŽgŽ, www.hec.ca/pages/simon.van-norden Service de l'enseignement de la finance, ƒcole des H.E.C. 3000 Cote-Sainte-Catherine, Montreal QC, CANADA H3T 2A7 simon.van-norden@hec.ca or (514)340-6781 or fax:(514)340-5632 From ???@??? Fri Jan 28 10:10:24 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id KAA19888 for ; Fri, 28 Jan 2000 10:05:04 +1100 Message-Id: <200001272305.KAA19888@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 28 Jan 2000 10:05:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 28 Jan 100 10:01:28 GMT+1000 From: "Estima" To: "RATS Discussion List" Subject: Re: Postscript Output of Graphs Date: Thu, 27 Jan 2000 16:58:03 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Estima MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v3.11) (via Mercury MTS (Bindery) v1.40) > I'm using RATS graphics for the first time in a long time and I'm having trouble > getting properly sized Postscript output. > > I want to place a stacked graph I created with SPGRAPH on a page in Portrait > orientation. When I print direct from GRAFEDIT, the bottom third of the page is > left blank. When I export from GRAFEDIT to postscript (Portrait), I get the top > third of the page left blank. Assuming you are using the Windows version, you have several choices: First, you have several ways to print the graph directly to the printer: 1) You can print graphs directly to your PostScript printer, in any of 4 layouts/orientations: A) If you use the default NOPORTRAIT setting on GRPARM, you can use the "Portrait" and "Landscape" choices in the Page Setup dialog box to choose between the two layouts shown on page 10 of the Version 4.2 manual supplement. B) If you use GRPARM(PORTRAIT) before generating your graphs, you'll have two other orientatios to choose from, depending on the "Portrait" or "Landscape" setting in the dialog box (the manual supplement suggests that GRPARM(PORTRAIT) does not work in Windows-- that is not correct, as we decided to add support for that shortly before 4.2 was released). You can also export the graph to a PostScript file. IMPORTANT NOTE: The exported graphs are optimized for placing into a word processing or page layout program, where you can position and re- size the graphs as desired before printing. They are not optimized for being dumped directly to a printer--you can do so, but the graph will be small and positioned right on one corner of the page. If you just want to print the graph directly to a PS printer, do it directly from RATS, or by using the Print command in GRAFEDIT, rather than converting the files. 1) You can export the graph directly to PostScript from RATS in either of 2 orientations by using File-Save As... and selecting either the "Portrait" or "Landscape" PostScript filter (using GRPARM(PORTRAIT) before generating the graph adds two other possible orientations). 2) If you prefer to use GRAFEDIT, and you have version 4.2 or later, you can still convert files to PS using either "Landscape" or "Portrait" by using the choices on the Xport menu. I believe the command line /PS switch only does the "Portrait" mode export. Remember, don't export the graph to a file if you just want to dump it to the printer--print directly from RATS instead. Sincerey, Tom Maycock Estima -- ------------------------------------------------------------ | Estima | Sales: (800) 822-8038 | | P.O. Box 1818 | (847) 864-8772 | | Evanston, IL 60204-1818 | Support: (847) 864-1910 | | USA | Fax: (847) 864-6221 | | http://www.estima.com | estima@estima.com | ------------------------------------------------------------ From ???@??? Fri Jan 28 13:13:18 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id NAA02552 for ; Fri, 28 Jan 2000 13:05:04 +1100 Message-Id: <200001280205.NAA02552@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 28 Jan 2000 13:05:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 28 Jan 100 13:03:56 GMT+1000 From: "Guido Travaglini" To: "RATS Discussion List" Subject: DOFOR & SET INSTRUCTIONS Date: Thu, 27 Jan 2000 18:03:25 PST Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; format=flowed X-Mailer: Mercury MTS (Bindery) v1.40 I've lately had problems with the SET command within the DOFOR loop. Let y=[x1...xn]' be the vector of variables with T=end observations and the loop be: dofor y = x1 ... xn lin(noprint) y start+5 end #constant y{1 to lag} ... set v start end = y**2 print start end y v acy end dofor As for LIN everything OK, but when SETting the new variables v, instead of getting y**2, I get printed the code numbers (squared) of the variables y. How can I DECLARE the index y or what else should be done to obtain correct results? Any help is appreciated. Thanx in advance, Guido. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From ???@??? Fri Jan 28 13:33:17 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id NAA24910 for ; Fri, 28 Jan 2000 13:25:03 +1100 Message-Id: <200001280225.NAA24910@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 28 Jan 2000 13:25:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 28 Jan 100 13:20:25 GMT+1000 From: Rob Trevor To: "RATS Discussion List" Subject: Re: DOFOR & SET INSTRUCTIONS Date: Fri, 28 Jan 2000 13:20:42 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" ; format="flowed" X-Mailer: Mercury MTS (Bindery) v1.40 At 6:03 PM -0800 27/1/00, Guido Travaglini wrote: >I've lately had problems with the SET command within the DOFOR >loop. Let y=[x1...xn]' be the vector of variables with T=end >observations and the loop be: > dofor y = x1 ... xn > lin(noprint) y start+5 end > #constant y{1 to lag} ... > set v start end = y**2 > print start end y v acy > end dofor >As for LIN everything OK, but when SETting the new variables >v, instead of getting y**2, I get printed the code numbers >(squared) of the variables y. How can I DECLARE the index y or >what else should be done to obtain correct results? Any help is >appreciated. Thanx in advance, Guido. RATS is getting confused between the INTEGER 'y' used in the 'dofor' loop and the integer 'y' used to differentiate between your series :) (This is a very common gotya.) From memory, something like replacing 'y**2' by 'y(t)**2' or declaring [SERIES] in the dofor instructions should work. See the manual chapter on programming for more info. (I did say 'from memory'...I'd check the manual if I were you!) Cheers Rob From ???@??? Fri Jan 28 13:53:06 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id NAA07620 for ; Fri, 28 Jan 2000 13:45:04 +1100 Message-Id: <200001280245.NAA07620@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 28 Jan 2000 13:45:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 28 Jan 100 13:43:40 GMT+1000 From: Peter Summers To: "RATS Discussion List" Subject: Re: DOFOR & SET INSTRUCTIONS Date: Fri, 28 Jan 2000 13:47:28 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.2 (via Mercury MTS (Bindery) v1.40) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Try set v start end = ([series]y)**2 or Rob Trevor's suggestion: y(t)**2 Pete At 18:03 27/01/00 -0800, you wrote: >I've lately had problems with the SET command within the DOFOR >loop. Let y=[x1...xn]' be the vector of variables with T=end >observations and the loop be: > dofor y = x1 ... xn > lin(noprint) y start+5 end > #constant y{1 to lag} ... > set v start end = y**2 > print start end y v acy > end dofor >As for LIN everything OK, but when SETting the new variables >v, instead of getting y**2, I get printed the code numbers >(squared) of the variables y. How can I DECLARE the index y or >what else should be done to obtain correct results? Any help is >appreciated. Thanx in advance, Guido. > > >______________________________________________________ >Get Your Private, Free Email at http://www.hotmail.com From ???@??? Fri Jan 28 19:48:12 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id TAA02486 for ; Fri, 28 Jan 2000 19:45:04 +1100 Message-Id: <200001280845.TAA02486@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 28 Jan 2000 19:45:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 28 Jan 100 19:40:22 GMT+1000 From: "Konstantinos Tsatsaronis" To: "RATS Discussion List" Subject: Re: DOFOR & SET INSTRUCTIONS Date: Fri, 28 Jan 2000 09:39:19 +0100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: X-Mailer: Novell GroupWise 5.5 (via Mercury MTS (Bindery) v1.40) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit This is one of the best examples of RATS capriciousness that irritates people. The index you use in the do loop is an integer and the set command requires a series. I have used two alternative ways to have the program understand what it is supposed to do: 1) you can declare the index in the DOFOR statement as a series DOFOR [series] i=x1... xn 2) Alternatively you should specify that the integer identifies a series in the commands that follow SET y = ([series] i)**2 3) Also, occasionally, if you also specify a lag (even if you don't need one) the command will understand what you want SET y = (i{0})**2 The problem is that for reasons that are not immediately obvious to me these tricks work or do not work in different programs I write. I never trust the result until I check it. The absence of error messages does not mean that it did what you intended it to do (like in all programs). Of course all of this is hidden in different places in the manual but unless you memorise the whole document you never get a good overview by looking at one place. Kostas Tsatsaronis >>> "Guido Travaglini" 01/28/00 03:03am >>> I've lately had problems with the SET command within the DOFOR loop. Let y=[x1...xn]' be the vector of variables with T=end observations and the loop be: dofor y = x1 ... xn lin(noprint) y start+5 end #constant y{1 to lag} ... set v start end = y**2 print start end y v acy end dofor As for LIN everything OK, but when SETting the new variables v, instead of getting y**2, I get printed the code numbers (squared) of the variables y. How can I DECLARE the index y or what else should be done to obtain correct results? Any help is appreciated. Thanx in advance, Guido. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com DISCLAIMER: Any e-mail messages from the Bank for International Settlements are sent in good faith, but shall not be binding nor construed as constituting any obligation on the part of the Bank. CONFIDENTIALITY NOTICE: This e-mail contains confidential information, which is intended only for the use of the recipient(s) named above. If you have received this communication in error, please notify the sender immediately via e-mail and return the entire message. Thank you for your assistance. From ???@??? Fri Jan 28 19:58:11 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id TAA08134 for ; Fri, 28 Jan 2000 19:50:04 +1100 Message-Id: <200001280850.TAA08134@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Fri, 28 Jan 2000 19:50:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 28 Jan 100 19:47:38 GMT+1000 From: "Konstantinos Tsatsaronis" To: "RATS Discussion List" Subject: Re: Postscript Output of Graphs Date: Fri, 28 Jan 2000 09:45:48 +0100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: X-Mailer: Novell GroupWise 5.5 (via Mercury MTS (Bindery) v1.40) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit I think that Simon's complaint was that even if you just print the graph directly from RATS you never get it to cover the whole page. It always seems to leave a lot of unused space. This is particularly problematic if you plot more than two graphs on one page (SPGRAPH). Kostas Tsatsaronis >>> "Estima" 01/27/00 11:58pm >>> > I'm using RATS graphics for the first time in a long time and I'm having trouble > getting properly sized Postscript output. > > I want to place a stacked graph I created with SPGRAPH on a page in Portrait > orientation. When I print direct from GRAFEDIT, the bottom third of the page is > left blank. When I export from GRAFEDIT to postscript (Portrait), I get the top > third of the page left blank. Assuming you are using the Windows version, you have several choices: First, you have several ways to print the graph directly to the printer: 1) You can print graphs directly to your PostScript printer, in any of 4 layouts/orientations: A) If you use the default NOPORTRAIT setting on GRPARM, you can use the "Portrait" and "Landscape" choices in the Page Setup dialog box to choose between the two layouts shown on page 10 of the Version 4.2 manual supplement. B) If you use GRPARM(PORTRAIT) before generating your graphs, you'll have two other orientatios to choose from, depending on the "Portrait" or "Landscape" setting in the dialog box (the manual supplement suggests that GRPARM(PORTRAIT) does not work in Windows-- that is not correct, as we decided to add support for that shortly before 4.2 was released). You can also export the graph to a PostScript file. IMPORTANT NOTE: The exported graphs are optimized for placing into a word processing or page layout program, where you can position and re- size the graphs as desired before printing. They are not optimized for being dumped directly to a printer--you can do so, but the graph will be small and positioned right on one corner of the page. If you just want to print the graph directly to a PS printer, do it directly from RATS, or by using the Print command in GRAFEDIT, rather than converting the files. 1) You can export the graph directly to PostScript from RATS in either of 2 orientations by using File-Save As... and selecting either the "Portrait" or "Landscape" PostScript filter (using GRPARM(PORTRAIT) before generating the graph adds two other possible orientations). 2) If you prefer to use GRAFEDIT, and you have version 4.2 or later, you can still convert files to PS using either "Landscape" or "Portrait" by using the choices on the Xport menu. I believe the command line /PS switch only does the "Portrait" mode export. Remember, don't export the graph to a file if you just want to dump it to the printer--print directly from RATS instead. Sincerey, Tom Maycock Estima -- ------------------------------------------------------------ | Estima | Sales: (800) 822-8038 | | P.O. Box 1818 | (847) 864-8772 | | Evanston, IL 60204-1818 | Support: (847) 864-1910 | | USA | Fax: (847) 864-6221 | | http://www.estima.com | estima@estima.com | ------------------------------------------------------------ DISCLAIMER: Any e-mail messages from the Bank for International Settlements are sent in good faith, but shall not be binding nor construed as constituting any obligation on the part of the Bank. CONFIDENTIALITY NOTICE: This e-mail contains confidential information, which is intended only for the use of the recipient(s) named above. If you have received this communication in error, please notify the sender immediately via e-mail and return the entire message. Thank you for your assistance. From ???@??? Sat Jan 29 13:07:30 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id BAA08364 for ; Sat, 29 Jan 2000 01:30:04 +1100 Message-Id: <200001281430.BAA08364@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 29 Jan 2000 01:30:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 29 Jan 100 01:09:34 GMT+1000 From: Simon van Norden To: "RATS Discussion List" Subject: Re: Postscript Output of Graphs Date: Fri, 28 Jan 2000 09:13:21 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Ecole des Hautes Etudes Commerciales X-Mailer: Mozilla 4.61 [en] (Win98; U) (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit I think I've got the problem pretty much licked (or at least, much improved.) I was using SPGRAPH to stack 3 graphs vertically on a page. I then wanted to produce output for postscript (actually, this was for conversion into Adobe Acrobat, but that's much the same.) Naturally, this looks much better in a PORTRAIT than in a LANDSCAPE orientation. I get pretty good results if I do the following. 1) As Tom Maycock suggested, use GRPARM(PORTRAIT). This makes the graphs look sideways on the screen. Note that this option does not seem to be documented in the on-line help in my version (4.30). 2) Use WinRATS (not GRAFEDIT) to print the graph. I did not try using GRAFEDIT with the GRPARM(PORTRAIT), but previous efforts showed that using WinRATS to display and print new graphs or saved RGF files gave better sized results. 3) Despite the fact that you want the graph in PORTRAIT orientation, choose LANDSCAPE as your printer's setting. This gives you a sideways graph on a sideways page, but the graph fills the page nicely. Konstantinos Tsatsaronis wrote: > > I think that Simon's complaint was that even if you just print the graph directly from RATS you never get it to cover the whole page. It always seems to leave a lot of unused space. This is particularly problematic if you plot more than two graphs on one page (SPGRAPH). > > Kostas Tsatsaronis > > >>> "Estima" 01/27/00 11:58pm >>> > > I'm using RATS graphics for the first time in a long time and I'm having trouble > > getting properly sized Postscript output. > > > > I want to place a stacked graph I created with SPGRAPH on a page in Portrait > > orientation. When I print direct from GRAFEDIT, the bottom third of the page is > > left blank. When I export from GRAFEDIT to postscript (Portrait), I get the top > > third of the page left blank. > > Assuming you are using the Windows version, you have several choices: > > First, you have several ways to print the graph directly to the > printer: > > 1) You can print graphs directly to your PostScript printer, in any > of 4 layouts/orientations: > > A) If you use the default NOPORTRAIT setting on GRPARM, you can use > the "Portrait" and "Landscape" choices in the Page Setup dialog box > to choose between the two layouts shown on page 10 of the Version 4.2 > manual supplement. > > B) If you use GRPARM(PORTRAIT) before generating your graphs, you'll > have two other orientatios to choose from, depending on the > "Portrait" or "Landscape" setting in the dialog box (the manual > supplement suggests that GRPARM(PORTRAIT) does not work in Windows-- > that is not correct, as we decided to add support for that shortly > before 4.2 was released). > > You can also export the graph to a PostScript file. > > IMPORTANT NOTE: The exported graphs are optimized for placing into a > word processing or page layout program, where you can position and re- > size the graphs as desired before printing. They are not optimized > for being dumped directly to a printer--you can do so, but the graph > will be small and positioned right on one corner of the page. If you > just want to print the graph directly to a PS printer, do it directly > from RATS, or by using the Print command in GRAFEDIT, rather than > converting the files. > > 1) You can export the graph directly to PostScript from RATS in > either of 2 orientations by using File-Save As... and selecting > either the "Portrait" or "Landscape" PostScript filter (using > GRPARM(PORTRAIT) before generating the graph adds two other possible > orientations). > > 2) If you prefer to use GRAFEDIT, and you have version 4.2 or later, > you can still convert files to PS using either "Landscape" or > "Portrait" by using the choices on the Xport menu. I believe the > command line /PS switch only does the "Portrait" mode export. > > Remember, don't export the graph to a file if you just want to dump > it to the printer--print directly from RATS instead. > > Sincerey, > Tom Maycock > Estima > > -- > ------------------------------------------------------------ > | Estima | Sales: (800) 822-8038 | > | P.O. Box 1818 | (847) 864-8772 | > | Evanston, IL 60204-1818 | Support: (847) 864-1910 | > | USA | Fax: (847) 864-6221 | > | http://www.estima.com | estima@estima.com | > ------------------------------------------------------------ > > DISCLAIMER: Any e-mail messages from the Bank for International Settlements are sent in good faith, but shall not be binding nor construed as constituting any obligation on the part of the Bank. > > CONFIDENTIALITY NOTICE: This e-mail contains confidential information, which is intended only for the use of the recipient(s) named above. If you have received this communication in error, please notify the sender immediately via e-mail and return the entire message. Thank you for your assistance. -- Simon van Norden, Prof. agrŽgŽ, www.hec.ca/pages/simon.van-norden Service de l'enseignement de la finance, ƒcole des H.E.C. 3000 Cote-Sainte-Catherine, Montreal QC, CANADA H3T 2A7 simon.van-norden@hec.ca or (514)340-6781 or fax:(514)340-5632 From ???@??? Sat Jan 29 13:07:56 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id DAA17450 for ; Sat, 29 Jan 2000 03:00:04 +1100 Message-Id: <200001281600.DAA17450@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 29 Jan 2000 03:00:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 29 Jan 100 02:29:58 GMT+1000 From: "Estima" To: "RATS Discussion List" Subject: Re: Postscript Output of Graphs Date: Fri, 28 Jan 2000 09:29:41 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Estima MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v3.11) (via Mercury MTS (Bindery) v1.40) > I think that Simon's complaint was that even if you just print the > graph directly from RATS you never get it to cover the whole page. > It always seems to leave a lot of unused space. This is > particularly problematic if you plot more than two graphs on one > page (SPGRAPH). But, as I said, you can fill the whole page, using the "Landscape" switch and/or by doing GRPARM(PORTRAIT) first. Sincerely, Tom Maycock > -- ------------------------------------------------------------ | Estima | Sales: (800) 822-8038 | | P.O. Box 1818 | (847) 864-8772 | | Evanston, IL 60204-1818 | Support: (847) 864-1910 | | USA | Fax: (847) 864-6221 | | http://www.estima.com | estima@estima.com | ------------------------------------------------------------ From ???@??? Sat Jan 29 13:07:58 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id EAA24928 for ; Sat, 29 Jan 2000 04:00:04 +1100 Message-Id: <200001281700.EAA24928@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 29 Jan 2000 04:00:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 29 Jan 100 03:45:16 GMT+1000 From: "Estima" To: "RATS Discussion List" Subject: Re: DOFOR & SET INSTRUCTIONS Date: Fri, 28 Jan 2000 10:45:56 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: Estima MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v3.11) (via Mercury MTS (Bindery) v1.40) > This is one of the best examples of RATS capriciousness that > irritates people. Sorry to hear it irritates people. While the behavior may surprise some people, it is definitely not unpredictable or random. The first thing to remember is that every series in RATS has an integer number associated with it (normally based on the order in which the series was created in that program). This is actually an amazingly useful feature that makes many things possible and other things easier. The ability to use arrays of series in Version 4 has replaced the use of series numbers to some extent, as I think they are easier to handle, series numbers can still be very useful. Most of the rest of this is pretty much explained in earlier messages, but I thought I would try to do it a bit more formally: Suppose you have a list of series X1, X2, and X3. Unless the index variable of a DOFOR has already been declared as a series, a command like: dofor s = x1 x2 x3 will define "S" as an integer variable, which will take on the series numbers of X1, X2, and X3 in turn. Something like: linreg s # regressors will work exactly as expected. RATS knows the first parameter on LINREG has to be a series, so it accepts either a series name or any integer-valued variable or expression. In the case of an integer expression, RATS uses the series with that number. In this case, that would be the series number for X1 on the first trip through, X2 the second trip through, and so on. The same is true if S appears on the left hand side of a SET or FMRL instruction, the series list on PRINT or TABLE, or anywhere else RATS is explicitly expecting a series. However, suppose you do: set depvar = s The right hand side of a SET expression could be just about anything, so RATS has no way of knowing that you want to treat S as anything other than a constant integer value. As suggested by others, there are two basic ways to tell RATS to treat an integer variable/expression as a series when it is not clear from the context. To the best of my knowledge, these will _always_ work as long as you do them correctly--there's nothing "occassional" about it: 1) The most general is to use the [series] type modifier. Enclose the variable name (or integer expression) and the [series] tag inside parentheses. For example: set depvar = ([series] s) If you want to refer to a specific entry (or use the T subscript in a SET or FRML), put the second set of parens immediately after the close paren in the "([series]s)" expression: compute x5 = ([series] s)(5) set depvar = ([series] s)(1) set depvar = ([series] s)(t-1) etc. 2) In a SET or FRML expression, you can also use lag notation, which will be sufficient to tell RATS that you want to treat the integer variable/expression as a series: set depvar = s{0} (sets deptvar(t) = entry t of s) set depvar = s{1} (sets deptvar(t) = entry t-1 of s) Hope this is helpful. Sincerely, Tom Maycock Estima -- ------------------------------------------------------------ | Estima | Sales: (800) 822-8038 | | P.O. Box 1818 | (847) 864-8772 | | Evanston, IL 60204-1818 | Support: (847) 864-1910 | | USA | Fax: (847) 864-6221 | | http://www.estima.com | estima@estima.com | ------------------------------------------------------------ From ???@??? Sat Jan 29 13:08:07 2000 Received: from localhost (localhost [127.0.0.1]) by amethyst.efs.mq.edu.au (8.9.1/8.9.1) with ESMTP id FAA08376 for ; Sat, 29 Jan 2000 05:00:04 +1100 Message-Id: <200001281800.FAA08376@amethyst.efs.mq.edu.au> Received: from efs01.efs.mq.edu.au by localhost with POP3 (fetchmail-5.0.0) for robt@localhost (single-drop); Sat, 29 Jan 2000 05:00:04 +1100 (EETDT) Received: from SpoolDir by EFS01 (Mercury 1.40); 29 Jan 100 04:49:56 GMT+1000 From: "Klaus P. Fischer" To: "RATS Discussion List" Subject: Re: DOFOR & SET INSTRUCTIONS Date: Fri, 28 Jan 2000 12:48:42 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs01.efs.mq.edu.au X-listname: Organization: CREFA-Université Laval X-Mailer: Mozilla 4.7 [fr] (Win98; I) (via Mercury MTS (Bindery) v1.40) MIME-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 8bit This is certainly the most coherent explicit presentation of this sometimes rather "unclear" (but, of course, after such a fine explanation, clear) issue. It wouldn't be a waste to "cut and paste" this explanation into future versions of the manual in the chapter on Programmers Tools (3 y think?). It would have helped me to save some hours of pondering and fiddling at do-loops formulations and results. Thanks, KFischer Laval University Estima a Žcrit : > > This is one of the best examples of RATS capriciousness that > > irritates people. > > Sorry to hear it irritates people. While the behavior may surprise > some people, it is definitely not unpredictable or random. > > The first thing to remember is that every series in RATS has an > integer number associated with it (normally based on the order in > which the series was created in that program). This is actually an > amazingly useful feature that makes many things possible and other > things easier. > > The ability to use arrays of series in Version 4 has replaced the use > of series numbers to some extent, as I think they are easier to > handle, series numbers can still be very useful. > > Most of the rest of this is pretty much explained in earlier > messages, but I thought I would try to do it a bit more formally: > > Suppose you have a list of series X1, X2, and X3. Unless the index > variable of a DOFOR has already been declared as a series, a command > like: > > dofor s = x1 x2 x3 > > will define "S" as an integer variable, which will take on the series > numbers of X1, X2, and X3 in turn. > > Something like: > > linreg s > # regressors > > will work exactly as expected. RATS knows the first parameter on > LINREG has to be a series, so it accepts either a series name or any > integer-valued variable or expression. In the case of an integer > expression, RATS uses the series with that number. In this case, that > would be the series number for X1 on the first trip through, X2 the > second trip through, and so on. > > The same is true if S appears on the left hand side of a SET or FMRL > instruction, the series list on PRINT or TABLE, or anywhere else RATS > is explicitly expecting a series. > > However, suppose you do: > > set depvar = s > > The right hand side of a SET expression could be just about anything, > so RATS has no way of knowing that you want to treat S as anything > other than a constant integer value. > > As suggested by others, there are two basic ways to tell RATS to > treat an integer variable/expression as a series when it is not clear > from the context. To the best of my knowledge, these will _always_ > work as long as you do them correctly--there's nothing "occassional" > about it: > > 1) The most general is to use the [series] type modifier. Enclose the > variable name (or integer expression) and the [series] tag inside > parentheses. For example: > > set depvar = ([series] s) > > If you want to refer to a specific entry (or use the T subscript in a > SET or FRML), put the second set of parens immediately after the > close paren in the "([series]s)" expression: > > compute x5 = ([series] s)(5) > set depvar = ([series] s)(1) > set depvar = ([series] s)(t-1) > etc. > > 2) In a SET or FRML expression, you can also use lag notation, which > will be sufficient to tell RATS that you want to treat the integer > variable/expression as a series: > > set depvar = s{0} (sets deptvar(t) = entry t of s) > set depvar = s{1} (sets deptvar(t) = entry t-1 of s) > > Hope this is helpful. > > Sincerely, > Tom Maycock > Estima > > -- > ------------------------------------------------------------ > | Estima | Sales: (800) 822-8038 | > | P.O. Box 1818 | (847) 864-8772 | > | Evanston, IL 60204-1818 | Support: (847) 864-1910 | > | USA | Fax: (847) 864-6221 | > | http://www.estima.com | estima@estima.com | > ------------------------------------------------------------