From: "Estima" To: "RATS Discussion List" Subject: Announcements from Estima Date: Fri, 31 Jan 1997 10:39:10 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v2.42) (via Mercury MTS (Bindery) v1.30) Dear RATS Users: I've recently posted several items that may be of interest on our Web site. I thought I'd summarize them here briefly. See http://www.estima.com for details. Also we'll be sending out a newsletter soon, so if you've moved recently, be sure to send us your new address. * RATS Version 4.21 is now shipping for DOS/Windows platforms. This is a minor bug-fix release--no new functionality. Maintenance contract users and users who helped identify/solve problems will be receiving copies automatically. Other 4.20 users can get the update for a handling fee of $5.00 (US customers) or $10.0 (outside the US). * RATS Version 4.30 is now shipping for UNIX systems, and will be the next update shipped (probably in a couple of months) for all other platforms. For UNIX, this adds the ability view graphs directly on the screen under X Windows. For all platforms, this adds neural net capabilities, built-in linear/quadratic programming, and more. * Several very useful new procs have been added to the Web site, including a menu-driven VAR proc, a GARCH procedure, and more. Also, an new I(2) add-on for CATS is available free of charge. * Walter Enders has written a RATS-based study-guide companion to his popular "Applied Econometric Time Series" text. It is called "RATS Handbook for Econometric Time Series" and is available now from Wiley publishing. If all goes well, you will soon be able to purchase it directly from us. Watch for details in the upcoming newsletter. Thanks, Tom Maycock Estima -- +-----------------------------+-----------------------------------------+ | Estima | | | P.O. Box 1818 | Voice: (847) 864-8772 | | Evanston, IL 60204-1818 | Fax: (847) 864-6221 | | U.S.A | BBS: (847) 864-8816 | | e-mail: estima@estima.com | CompuServe: 73140,2202 | |-----------------------------------------------------------------------| | Web Site: http://www.estima.com | | RATS Internet Mailing List: New members can join by sending e-mail to | | MAISER@EFS.MQ.EDU.AU with the message: SUBSCRIBE RATS-L | +-----------------------------------------------------------------------+ ---------- End of message ---------- From: WHEELER@wmich.edu To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Sun, 02 Feb 1997 18:11:27 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT X-Mailer: Mercury MTS (Bindery) v1.30 Hi Tom, I have a question. When I estimate a VAR with "estimate", RATS allows me to save the coefficents in a number series. I would like to estimate a VAR, run the coefficents through modify and vreplace, and save the modified coefficients in a number series. I know how to use modify and vreplace, but how do I save the "new" coefficents in a numbered series. Mark Wheeler ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 3 Feb 1997 11:48:35 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 At 3:39 AM +1100 1/2/97, Estima wrote: > ... > * RATS Version 4.30 is now shipping for UNIX systems, and will be > the next update shipped (probably in a couple of months) for all > other platforms. For UNIX, this adds the ability view graphs directly > ... I really hope that this time the Mac version is NOT vapourware! Keep your heads down and get it out, PLEASE! Rob Trevor ---------- End of message ---------- From: Gregory Lypny To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 3 Feb 97 08:50:57 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: x-mailer: Claris Emailer 1.1 (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Everyone sing please (to the tune of the Beatle's "All Together Now"), Where's the MAC now? Where's the MAC now? Where's the MAC now? Where's the MAC now? Where's the MAC now? Where's the MAC now? Where's the MAC now? Where's the MAC now? Regards, Greg Gregory J. Lypny Associate Professor of Finance Director, Centre for Instructional Technology Mailing address: Contact: ------------------------------------------------------------------- Finance Department 514.848.2926 (voice) Concordia University 514.848.4500 (fax) 1455 De Maisonneuve Boulevard West LYPNY@VAX2.CONCORDIA.CA Montreal, QC H3G 1M8 ---------- End of message ---------- From: Per Homer To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 3 Feb 1997 09:10:17 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: Bank of Canada X-Mailer: Mozilla 3.01 (X11; I; SunOS 5.5.1 sun4u) (via Mercury MTS (Bindery) v1.30) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I followed the unsubscribe protocol given in the last email, but I still receiving mail from your list. Please unsubscribe me now. Thanks, Per Homer mailto:phomer@bank-banque-canada.ca ---------- End of message ---------- From: Mason Gerety To: "RATS Discussion List" Subject: Announcements from Estima -Reply Date: Mon, 03 Feb 1997 08:44:59 -0700 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Novell GroupWise 4.1 (via Mercury MTS (Bindery) v1.30) Please unsubsribe me. ---------- End of message ---------- From: "Estima" To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 3 Feb 1997 11:32:16 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v2.42) (via Mercury MTS (Bindery) v1.30) > Everyone sing please (to the tune of the Beatle's "All Together Now"), > > Where's the MAC now? I will say that I have personally witnessed actual work being done on the Mac version of RATS :-). Also, I'd be interested to hear your opinions on Apple's recent operating system two-step. What are the future plans of MacRATS users regarding their systems: 1. Continue running MacOS/Mac hardware indefinitely. 2. Stick with Apple, and upgrade to the "NeXT" version of the OS (to be called "Rhapsody" I believe). 3. Wait and see on Rhapsody while sticking with MacOs. 4. Consider switching to another platform. 5. Something else? Let me know. Tom Maycock Estima -- +-----------------------------+-----------------------------------------+ | Estima | | | P.O. Box 1818 | Voice: (847) 864-8772 | | Evanston, IL 60204-1818 | Fax: (847) 864-6221 | | U.S.A | BBS: (847) 864-8816 | | e-mail: estima@estima.com | CompuServe: 73140,2202 | |-----------------------------------------------------------------------| | Web Site: http://www.estima.com | | RATS Internet Mailing List: New members can join by sending e-mail to | | MAISER@EFS.MQ.EDU.AU with the message: SUBSCRIBE RATS-L | +-----------------------------------------------------------------------+ ---------- End of message ---------- From: June Y Hoang To: "RATS Discussion List" Subject: Ranel Data Analysis Date: Mon, 3 Feb 1997 12:54:09 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Greetings. I have been looking for any Rats Programs dealing with Dummy Variables Models and Random Effects Models (Panel Data Analysis). I would appreciate it if you could let me know where I could find them. Thanks for your time, June ---------- End of message ---------- From: "Christopher F. Baum" To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 03 Feb 1997 13:04:07 -0400 (EDT) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 I will stick with Apple, and when it appears that Rhapsody is a better environment than the MacOS, I'll use it. I wouldn't consider switching to a Wintel platform under any circumstances; however I do more and more of my work in a UNIX environment (including RATS; I can't effectively use the existing version on a PowerMac), so if Apple's directions are toward a UNIX-based system, that's just fine with me. UNIX beats WinNT any day of the week! Kit Baum baum@bc.edu ---------- End of message ---------- From: "Estima" To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 3 Feb 1997 13:19:46 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v2.42) (via Mercury MTS (Bindery) v1.30) > Hi Tom, I have a question. When I estimate a VAR with "estimate", RATS > allows me to save the coefficents in a number series. I would like to > estimate a VAR, run the coefficents through modify and vreplace, and save the > modified coefficients in a number series. I know how to use modify and > vreplace, but how do I save the "new" coefficents in a numbered series. Mark: The first thing that comes to mind would be to use the %EQNCOEFFS function to return an array containing the coefficients, and then to copy the values from the array into a series. The example below uses SCRATCH to create some numbered series. It might be easier to use a VECTOR of SERIES instead, but either should work. * VAR setup here. Now estimate: estimate * * Change equations as needed here. * * NUMEQ and NUMCOEFFS store number of equations and * number of coefficients per equation: com numeq = 2 com numcoeffs = 5 scratch numeq*numcoeffs / base compute counter = base do eqnnum=1,numeq compute counter = counter+1 compute coeffs = %eqncoeffs(eqnnum) set counter 1 numcoeffs = coeffs(t,1) end do print / base+1 base+2 Let me know if you have any questions. Sincerely, Tom Maycock Estima -- +-----------------------------+-----------------------------------------+ | Estima | | | P.O. Box 1818 | Voice: (847) 864-8772 | | Evanston, IL 60204-1818 | Fax: (847) 864-6221 | | U.S.A | BBS: (847) 864-8816 | | e-mail: estima@estima.com | CompuServe: 73140,2202 | |-----------------------------------------------------------------------| | Web Site: http://www.estima.com | | RATS Internet Mailing List: New members can join by sending e-mail to | | MAISER@EFS.MQ.EDU.AU with the message: SUBSCRIBE RATS-L | +-----------------------------------------------------------------------+ ---------- End of message ---------- From: WHEELER@wmich.edu To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 03 Feb 1997 14:13:47 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT X-Mailer: Mercury MTS (Bindery) v1.30 ---------- End of message ---------- From: WHEELER@wmich.edu To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 03 Feb 1997 14:14:14 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT X-Mailer: Mercury MTS (Bindery) v1.30 Thanks Tom. ---------- End of message ---------- From: Gregory Lypny To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Mon, 3 Feb 97 14:26:10 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: x-mailer: Claris Emailer 1.1 (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" >2. Stick with Apple, and upgrade to the "NeXT" version of the OS (to >be called "Rhapsody" I believe). > >3. Wait and see on Rhapsody while sticking with MacOs. 2 or 3 above. Greg Gregory J. Lypny Associate Professor of Finance Director, Centre for Instructional Technology Mailing address: Contact: ------------------------------------------------------------------- Finance Department 514.848.2926 (voice) Concordia University 514.848.4500 (fax) 1455 De Maisonneuve Boulevard West LYPNY@VAX2.CONCORDIA.CA Montreal, QC H3G 1M8 ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Thu, 6 Feb 1997 10:33:59 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 At 4:32 AM +1100 4/2/97, Estima wrote: > ... > I will say that I have personally witnessed actual work being done > on the Mac version of RATS :-). I hope that was in 1997 and not 1996 or even 1995 :-) [Yes, it has been that long!] > Also, I'd be interested to hear your opinions on Apple's recent > operating system two-step. What are the future plans of MacRATS users > regarding their systems: > > 1. Continue running MacOS/Mac hardware indefinitely. Yes, until its no longer able to efficiently perform the jobs it needs to do > 2. Stick with Apple, and upgrade to the "NeXT" version of the OS (to > be called "Rhapsody" I believe). Yes I will upgrade OS, provided backward compatibility with existing apps is as seems to be intended. > 3. Wait and see on Rhapsody while sticking with MacOs. I'll switch to new OS as soon as I can see that it provides stable backward compatibility > 4. Consider switching to another platform. I'd be open to this, if there was another platform which provided the same efficient, flexible working environment. This is not a simple one-eyed statement. At work I have 2 Mac's, 4 RS6000's and 4 Dell Pentium's. At home I have 2 Mac's. I love the Mac's for their interface and ease of moving from one program to another. I love the RS6000's for their power, and ease of admin via SMIT. The Pentium's I have for my RA's, because they have been trained to believe that you need a PC. However, they are always amazed at how easy it is for me to do tasks that they find extremely difficult on a PC. I'll give two quick examples. For programmers, have a look at BBEdit on the Mac (the full commercial version) and then find me a Windows text editor with the same power. (I've been looking for one for my RA's and simply can't find anything anywhere near BBEdit.) For RATS users, simply compare the ease of the Mac interface with the distinction between the RETURN and ENTER keys - hit return to end lines while editing, hit enter to execute selected lines - with the unavoidable messing around of the Windows version's READY/LOCAL modes. The Mac allows you to concentrate on the real problem, not the computer/software interface. While I can get the software and hardware solutions I need in the Mac world, I will use Mac's. Unless, of course, something miraculous occurs and Microsoft actual produce an OS that actually does allow you to concentrate on the real problem. I won't hold my breath, because the economics suggest that its in MS's interests to stimulate demand for that enormous consultant/support/book industry that interfaces between MS and the end-user. > 5. Something else? > > Let me know. > > Tom Maycock > Estima ---------- End of message ---------- From: RAY DEGENNARO To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Wed, 05 Feb 97 21:33:19 LCL Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Mercury MTS (Bindery) v1.30 On Thu, 6 Feb 1997 10:33:59 +1100 Rob Trevor said: >At 4:32 AM +1100 4/2/97, Estima wrote: >> > ... ... > >> 4. Consider switching to another platform. > >I'd be open to this, if there was another platform which provided the same >efficient, flexible working environment. > Try running RATS in a DOS window under OS/2. Granted, I'm a zealot, but I define Mac users as people who didn't compare it with OS/2. I have run 5-6 RATS jobs in the background while I did word processing in the foreground. If I somehow managed to crash one job, the others ran to completion despite it. By the way, you'll have full access to any editor for your RATS programs while all of this transpires. Ray DeGennaro ---------- End of message ---------- From: Paul Johnson To: "RATS Discussion List" Subject: RE: Announcements from Estima Date: Thu, 06 Feb 1997 10:34:52 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Content-transfer-encoding: 7BIT X-Mailer: Mercury MTS (Bindery) v1.30 >I'll give two quick examples. For programmers, have a look at BBEdit on the >Mac (the full commercial version) and then find me a Windows text editor >with the same power. (I've been looking for one for my RA's and simply >can't find anything anywhere near BBEdit.) I've never tried BBEdit but I have found "Programmer's File Editor" quite good. It is available free of charge from http://www.lancs.ac.uk/people/cpaap/pfe/ ---------- End of message ---------- From: "Jan bartholdy" To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Fri, 07 Feb 97 06:49:34 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: PMMail 1.53 For OS/2 UNREGISTERED SHAREWARE (via Mercury MTS (Bindery) v1.30) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit On Wed, 05 Feb 97 21:33:19 LCL, RAY DEGENNARO wrote: >> >Try running RATS in a DOS window under OS/2. I have been running RATS under OS/2 for about 4 years (since version 2.0) in fact I changed to OS/2 to be able to run RATS/SAS and fortran programs in the background. It is great! Now my question will Estima bring out an OS/2 version ? jan ----------------------------------------------------------------------------------------------------------------------------------------- Jan Bartholdy, Ph.D. University of Otago Department of Finance and Quantitative Analysis P.O. Box 56 Dunedin New Zealand (03) 479-8072 Fax: (03) 479-8193 E:mail: jbarthol@commerce.otago.ac.nz ---------- End of message ---------- From: ladslous mwansa To: "RATS Discussion List" Subject: HELP WITH VAR.SRC Date: Thu, 06 Feb 1997 19:25:31 -0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: handels X-Mailer: Mozilla 3.0 (Win16; I) (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi Everyone, I am trying to use VAR.SRC with RATS 4.20, but I don't seem to do it right. Can anyone who has tried this procedure help me with how to use this procedure? Ladslous ---------- End of message ---------- From: "Estima" To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Thu, 6 Feb 1997 13:09:10 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v2.42) (via Mercury MTS (Bindery) v1.30) > It is great! Now my question will Estima bring out an OS/2 > version ? Dear Jan et al: Probably not. We've asked for feedback on this in newletters a couple of times in the past, and gotten reponses from about 5 people each time--hardly enough to justify the porting work required, particularly since the DOS and Windows version work pretty well under OS/2. I just don't think there's a market for it. I have little experience with OS/2 myself, but from what I do know, it may indeed be the best OS around for personal computers. Unfortunately, IBM has bungled their operating system policy even worse than Apple (no mean feat that). OS/2 for the PowerPC could have been a killer product, particularly if they could have provided emulation for Wintel apps. It boggles my mind that IBM would fork over the money to help develop the PowerPC chip and then not come up with anything other than a port of AIX. Oh well. Lately, I've gotten so tired of hearing conflicting or flip/flopping stories from IBM that I've lost track of their current promises about OS/2--I assume that OS/2 for PowerPC is still a dead issue? Anyway, thanks to Mac users for their thoughts on future operating systems. I hope there's more of you out there we haven't heard from yet--speak up if there are. By the way, I certainly didn't/don't mean to start a PC vs. Mac debate. Although I think those are always good for a few chuckles, there are probably better forums for that. Thanks, Tom Maycock Estima -- +-----------------------------+-----------------------------------------+ | Estima | | | P.O. Box 1818 | Voice: (847) 864-8772 | | Evanston, IL 60204-1818 | Fax: (847) 864-6221 | | U.S.A | BBS: (847) 864-8816 | | e-mail: estima@estima.com | CompuServe: 73140,2202 | |-----------------------------------------------------------------------| | Web Site: http://www.estima.com | | RATS Internet Mailing List: New members can join by sending e-mail to | | MAISER@EFS.MQ.EDU.AU with the message: SUBSCRIBE RATS-L | +-----------------------------------------------------------------------+ From: Rob Trevor To: "RATS Discussion List" Subject: ADMIN: Unreachable Subscribers Date: Fri, 7 Feb 1997 10:06:08 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 Sorry about the interuption, but if you happen to recognize any of the following email address and know how to contact the person, could you please email me directly? wlee@pop.hbs.edu kcyree@utkux.utcc.utk.edu ringenito@frbsfdoe.sanf.frb.org frbsfdoe!ringenito@frbsfdoe.sanf.frb.org pc1505@wow.com ilumb@CCMAIL-UK.PWCM.COM serfgrb@mpsenet.com.mx frbkc!pcooper@uunet.uu.net THIERRY.MICHEL@prevision.finances.fr DON5437@chollian.dacom.co.kr Luis.Portugal@individual.puug.pt yasushi@magical.egg.or.jp kamal@cdedse.ernet.in Email to each of these addresses has consistently bounced over a number of months. I have been unable to establish contact any of these people, and will need to delete them from the RATS-L list unless I am able to contact them and resolve the bounce problems. Thanks Rob ---------- End of message ---------- From: Economic Group To: "RATS Discussion List" Subject: Re: Announcements from Estima Date: Fri, 7 Feb 1997 10:57:49 +1000 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 Tom, >Also, I'd be interested to hear your opinions on Apple's recent >operating system two-step. What are the future plans of MacRATS users >regarding their systems: > >1. Continue running MacOS/Mac hardware indefinitely. Yes > >2. Stick with Apple, and upgrade to the "NeXT" version of the OS (to >be called "Rhapsody" I believe). If Rhapsody lives up to expectations we will move over, otherwise we will stick to 7.x until the NeXT product is sorted out. > >3. Wait and see on Rhapsody while sticking with MacOs. Yes > >4. Consider switching to another platform. Maybe CHRP running Mac OS / Win NT. > >5. Something else? Mathematica for all ? Thanks, Mark Gibson :-{) Economic Analysis Department Email enquiries: Economic Group PH: +612 9551 8817 Reserve Bank of Australia FAX: +612 9551 8033 ---------- End of message ---------- From: Hung-Jen Wang To: "RATS Discussion List" Subject: compute Date: Tue, 11 Feb 1997 00:00:00 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Hi, How could I let RATS understand that X is 2.5 in the following example? It seems RATS takes X to be INTEGER. ======================= compute X = 5/2 display X 2 ======================= It doesn't help if I declare the type to be REAL. ========================= | compute [real] Y = 5/2 | display Y | 2.00000 | ======================== The same problem arises with any fraction expression. ============================== allocate 1 3 | set Z = %if(0.1 > 1/2, 1 ,0)| print / Z | | | ENTRY Z | 1 1.0000000000000 | 2 1.0000000000000 | 3 1.0000000000000 | ============================ Any suggestion will be appreciated! H.J. Wang ---------- End of message ---------- From: "Dimitris Hatzinikolaou" To: "RATS Discussion List" Subject: Re: compute Date: Tue, 11 Feb 1997 16:16:34 GMT-0930 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: Flinders University Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT X-Mailer: Pegasus Mail for Windows (v2.52) (via Mercury MTS (Bindery) v1.30) Try putting decimal points, e.g., 5.0/2.0. See RATS manual, version 4.2, Chapter 1, p. 31. Cheers Dimitris Hatzinikolaou, lecturer The Flinders University of South Australia School of Economics GPO Box 2100 Adelaide, SA 5001 AUSTRALIA ---------- End of message ---------- From: Hung-Jen Wang To: "RATS Discussion List" Subject: compute again Date: Tue, 11 Feb 1997 02:29:14 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Hi, I thank those who give me quick replies of my previous message and suggested me add decimal points. However, this may not be very useful in my case.... Let me try to explain the situation. What I tried to do is the following: ================ * After running a regression of Y = bX + e ..... make(transpose,lastreg) Xtransp ; * Get the X' matrix call it Xtransp. compute raw = %raws(Xtransp) compute col = %cols(Xtransp) compute criterion = 2*col/raw ; * HERE IS THE PROBLEM! compute diag=%mqformdiag(%XX,Xtransp) ;* Get the diagonal elements of the * projection matrix; call it diag. set Z = %if( diag(T) > criterion, %NA, diag(T) ) ======================== Since "raw" and "col" are just the numbers of raws and columns of a matrix, they must be integers! Thus when I computed the "criterion" constant, RATS thinks that "criterion = 2*col/raw" is an integer. Of course I can manually key in the numbers of raws and columns, but it requires tedious book keepings, and, in most of the cases, I wouldn't know the number until finish the regression (because some observations would be dropped due to missing values). I hate to divide a program into two! Again, I will appreciate any comment! Thanks! H.J. Wang ---------- End of message ---------- From: "David Edgerton" To: "RATS Discussion List" Subject: Re: compute again Date: Tue, 11 Feb 1997 09:39:22 +0000 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Windows (v2.10) (via Mercury MTS (Bindery) v1.30) The problem with integer division is a common one with my students. I simply tell them to always multiply numerators and/or denominators by 1.0, that is compute criterion = 2*(1.*col)/raw ; instead of compute criterion = 2*col/raw ; However, I feel RATS priorities are all wrong. The default should be floating point division. Integer division should be the exception, using some sort of fix(.) or intdiv(.) function. Sincerely David Edgerton ----------------------------------------------------------------- David Edgerton tel: (+46) 46 222 8678 Dept of Economics fax: (+46) 46 222 4613 P.O. Box 7082 S-220 07 Lund e-mail: David.Edgerton@nek.lu.se Sweden ---------- End of message ---------- From: "John Frain" To: "RATS Discussion List" Subject: Re: compute again Date: Tue, 11 Feb 1997 11:36:28 -0000 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Microsoft Internet Mail 4.70.1155 (via Mercury MTS (Bindery) v1.30) MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Use the float function as follows: compute x = float(3)/float(2) disp x 1.50000 i.e. compute raw = float(%raws(Xtransp)) John Frain Phone (+3531)6716666 Head of Research Fax (+3531)6716561 Economics (ERP) Department Central Bank of Ireland POB 559 Dame St. Dublin, 2 Ireland ---------- > From: Hung-Jen Wang > To: RATS Discussion List > Subject: compute again > Date: 11 February 1997 07:29 > > Hi, > > I thank those who give me quick replies of my previous message > and suggested me add decimal points. However, this may > not be very useful in my case.... Let me try to explain the situation. > > What I tried to do is the following: > > ================ > * After running a regression of Y = bX + e ..... > > > make(transpose,lastreg) Xtransp ; * Get the X' matrix call it Xtransp. > > compute raw = %raws(Xtransp) > compute col = %cols(Xtransp) > > compute criterion = 2*col/raw ; * HERE IS THE PROBLEM! > > compute diag=%mqformdiag(%XX,Xtransp) ;* Get the diagonal elements of the > * projection matrix; call it diag. > > set Z = %if( diag(T) > criterion, %NA, diag(T) ) > ======================== > > Since "raw" and "col" are just the numbers of raws and columns of > a matrix, they must be integers! Thus when I computed the "criterion" > constant, RATS thinks that "criterion = 2*col/raw" is an integer. > > Of course I can manually key in the numbers of raws and columns, > but it requires tedious book keepings, and, in most of the cases, I wouldn't > know the number until finish the regression (because some observations > would be dropped due to missing values). I hate to divide a program into > two! > > Again, I will appreciate any comment! Thanks! > > H.J. Wang ---------- End of message ---------- From: Janet Netz To: "RATS Discussion List" Subject: Mac O/S Date: Tue, 11 Feb 1997 08:04:19 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Is the discussion list archived somewhere? I want to get those messages from a week or two ago on Mac O/S and what people planned to do. Thanks netz@mgmt.purdue.edu ---------- End of message ---------- From: Michael Hanson To: "RATS Discussion List" Subject: Mac Comments (Re: Announcements from Estima) Date: Tue, 11 Feb 1997 10:02:37 -0500 (EST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 With regard to Tom Haycock's questions about current Mac users: Assuming RATS for the Macintosh is out in the near future, I foresee myself upgrading to that. However, in the absence of a stable program, I have learned to get along quite well with a combination of Matlab, TSP (PowerMac native(!), but also a bit unstable) and Mathematica. (I understand Wolfram does much of its development on PowerMacs.) I don't see myself switching platforms before I switch programs; my current hardware, other software, and human capital investment is too high. I have occasion to use a number of platforms (MacOS, Un*x, Windows), and personally I find the first to be preferable in many, many aspects. Since I do not foresee Microsoft closing the gap with Apple (yes, I've tried Windows95) nor expect Apple to bungle the movement to a Mach-based OS so badly that the gap is closed (indeed, I expect the gap to widen), I expect to be using the MacOS on the most cost-effective hardware (from Apple or a clone manufacturer) to run it, well into the next century. Assuming there will be a version of RATS which runs on my machine, I'll probably own that as well. ;) -- Mike ====================================================================== Michael S. Hanson mhanson@umich.edu Department of Economics http://www.econ.lsa.umich.edu/~mhanson University of Michigan ---------- End of message ---------- From: MATT SALOMON To: "RATS Discussion List" Subject: Try Date: Tue, 11 Feb 1997 10:46:28 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Novell GroupWise 4.1 (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain Try x = 5./2. or, more generally, compute i = 5 compute j = 2 x = float(i)/float(j) Good luck.. Matt Salomon >>> Hung-Jen Wang 02/11/97 12:00am >>> Hi, How could I let RATS understand that X is 2.5 in the following example? It seems RATS takes X to be INTEGER. ======================= compute X = 5/2 display X 2 ======================= It doesn't help if I declare the type to be REAL. ========================= | compute [real] Y = 5/2 | display Y | 2.00000 | ======================== The same problem arises with any fraction expression. ============================== allocate 1 3 | set Z = %if(0.1 > 1/2, 1 ,0)| print / Z | | | ENTRY Z | 1 1.0000000000000 | 2 1.0000000000000 | 3 1.0000000000000 | ============================ Any suggestion will be appreciated! H.J. Wang ---------- End of message ---------- From: MATT SALOMON To: "RATS Discussion List" Subject: compute criterion = 2*float(col)/float(raw) ; * <<<<<<< HERE IS Date: Tue, 11 Feb 1997 10:48:52 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Novell GroupWise 4.1 (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain compute criterion = 2*float(col)/float(raw) ; * <<<<<<< HERE IS THE SOLUTION >>> Hung-Jen Wang 02/11/97 02:29am >>> Hi, I thank those who give me quick replies of my previous message and suggested me add decimal points. However, this may not be very useful in my case.... Let me try to explain the situation. What I tried to do is the following: ================ * After running a regression of Y = bX + e ..... make(transpose,lastreg) Xtransp ; * Get the X' matrix call it Xtransp. compute raw = %raws(Xtransp) compute col = %cols(Xtransp) compute criterion = 2*col/raw ; * HERE IS THE PROBLEM! compute diag=%mqformdiag(%XX,Xtransp) ;* Get the diagonal elements of the * projection matrix; call it diag. set Z = %if( diag(T) > criterion, %NA, diag(T) ) ======================== Since "raw" and "col" are just the numbers of raws and columns of a matrix, they must be integers! Thus when I computed the "criterion" constant, RATS thinks that "criterion = 2*col/raw" is an integer. Of course I can manually key in the numbers of raws and columns, but it requires tedious book keepings, and, in most of the cases, I wouldn't know the number until finish the regression (because some observations would be dropped due to missing values). I hate to divide a program into two! Again, I will appreciate any comment! Thanks! H.J. Wang ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: Re: compute again Date: Wed, 12 Feb 1997 10:39:53 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 At 8:39 PM +1100 11/2/97, David Edgerton wrote: > ... > However, I feel RATS priorities are all wrong. The default should be > floating point division. Integer division should be the exception, > using some sort of fix(.) or intdiv(.) function. > ... Hmm, I can see your point, however I'm not sure that this wouldn't just confuse those with a programming background. Integers are more 'precise' than floating point numbers, and I'm not sure that it makes sense to default to throwing this 'precision' out when I do the equivalent of i/2 in a loop for example. I can't imagine why we would want integer *multiply* to default to floating point - so perhaps we should not want divide to default either. Rob ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: Archives [was Re: Mac O/S] Date: Wed, 12 Feb 1997 11:19:55 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 Janet and others At 12:04 AM +1100 12/2/97, Janet Netz wrote: > Is the discussion list archived somewhere? I want to get those messages > from a week or two ago on Mac O/S and what people planned to do. The list is automatically archived by the month (Sydney time). [The list started at the end of January 1996.] To retrieve the archives for January and February 1997 send the following email: To: MAISER@efs.mq.edu.au Subject: SEND RATS9701.ARC SEND RATS9702.ARC The 'SEND's need to be at the very beginning of the body of the email message, one per line. Cheers Rob ----------------------------------------------------------------------- Rob Trevor Associate Professor Centre for Studies in Money, Banking and Finance Macquarie University, NSW, AUSTRALIA 2109 Internet: robt@efs.mq.edu.au VOICE: +61 (2) 9850-8447 FAX: +61 (2) 9850-7281 or +61 (2) 9418-1847 ----------------------------------------------------------------------- ---------- End of message ---------- From: mihyung kim To: "RATS Discussion List" Subject: floating point error Date: Thu, 13 Feb 1997 04:47:21 -0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: Chunbuk Sanup University X-Mailer: Mozilla 3.01 (Win95; I; 16bit) (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Does anyone know how to solve the following problem? I get error message of "Floating point of square root of negative number" when I run MAXIMIZE procedure and then my PC does not work any more. I wil expect your wisdom. Sincerely, Mihyung Kim E-mail: mihyung@sunny.howon.ac.kr ---------- End of message ---------- From: "Deitch, Jonathan" To: "RATS Discussion List" Subject: cointegration: pi matrix of full rank Date: Wed, 12 Feb 97 11:12:00 EST Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Microsoft Mail V3.0 (via Mercury MTS (Bindery) v1.30) This is more of a substantive question than a technical one, but here goes... Using CATS (johansen ml procedure), I have tested for cointegration in a 5-variable system. According to the L-max and L-trace tests, the pi matrix is of *full rank* and my diagnostic tests indicate white noise residuals. It would appear that my next step to estimate a VAR in differences and then identify equations for innovation accounting? Can anyone confirm this for me? I can't find any examples in the literature of this happening (although I can understand why if the researcher is indeed attempting to discover a cointegrating relationship...) PS: Feel free to be as verbose as possible. This is my first attempt with the Johansen procedure and I could use all the advice I can get. (I already have Enders 1995) Thanks a million, Jonathan Deitch Pentagon Federal Credit Union Alexandria, VA USA email: b0deitch@hq.PenFed.org ---------- End of message ---------- From: "Christopher F. Baum" To: "RATS Discussion List" Subject: discussion list archives on the Web Date: Wed, 12 Feb 1997 17:25:01 -0400 (EDT) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Dear RATSelers, The RATS-L discussion list archives may be viewed on the Web from the URL http://fmwww.bc.edu/ec-p/ratsL.html This URL is not case sensitive (one of the advantages of Made with Macintosh!). --------------------------------------------------------------------- Christopher F Baum Boston College Econ, Chestnut Hill MA 02167 USA baum@bc.edu fax 617 552 2308 http://fmwww.bc.edu/EC-V/baum.fac.html ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: Re: discussion list archives on the Web Date: Thu, 13 Feb 1997 09:54:35 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 Chris > The RATS-L discussion list archives may be viewed on the Web from the > URL > > http://fmwww.bc.edu/ec-p/ratsL.html > ... Well done and many thanks. Perhaps Tom can add a link from Estima's home page? Cheers Rob ----------------------------------------------------------------------- Rob Trevor Associate Professor Centre for Studies in Money, Banking and Finance Macquarie University, NSW, AUSTRALIA 2109 Internet: robt@efs.mq.edu.au VOICE: +61 (2) 9850-8447 FAX: +61 (2) 9850-7281 or +61 (2) 9418-1847 ----------------------------------------------------------------------- ---------- End of message ---------- From: Dick Durevall To: "RATS Discussion List" Subject: Re: cointegration: pi matrix of full rank Date: Thu, 13 Feb 1997 11:50:18 +0100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Windows Eudora Light Version 1.5.4 (16) (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" If the pi matrix is of *full rank* then all variables are stationary, i.e., you have five cointegratring vectors, and you can estimate the VAR in levels. Dick At 11.12 1997-02-12 EST, you wrote: > >This is more of a substantive question than a technical one, but here >goes... > >Using CATS (johansen ml procedure), I have tested for cointegration in a >5-variable system. >According to the L-max and L-trace tests, the pi matrix is of *full rank* >and my diagnostic tests >indicate white noise residuals. > >It would appear that my next step to estimate a VAR in differences and >then >identify equations for innovation accounting? Can anyone confirm this >for me? I can't >find any examples in the literature of this happening (although I can >understand why if >the researcher is indeed attempting to discover a cointegrating >relationship...) > >PS: Feel free to be as verbose as possible. This is my first attempt >with the Johansen procedure >and I could use all the advice I can get. (I already have Enders 1995) > >Thanks a million, >Jonathan Deitch >Pentagon Federal Credit Union >Alexandria, VA USA > >email: b0deitch@hq.PenFed.org > > ---------- End of message ---------- From: "F.Y.Kumah" To: "RATS Discussion List" Subject: Re: cointegration: pi matrix of full rank Date: Thu, 13 Feb 1997 11:06:43 MET Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: Tilburg University X-mailer: Pegasus Mail for Windows (v2.23) (via Mercury MTS (Bindery) v1.30) Dear Jonathan and Dick, I think the issue of whether or not to estimate the VAR in levels or first differences depends on the nature of the empirical analysis involved. Specifically, if one is interested in high frequency fluctuations in financial variables to some identified shocks would it be informative to take first differences? Though most financial variables might be I(1). I think the use of first differences in VARs of this nature might take away vital infornmation. To solve this problem some people may suggest running two VARs - one in leves and the other in differences. But as economists I think the underlying decisive factor should be the nature of the investigation. Further, I think cointegration is only interesting and informative if one could test some economic hypotheses using the estimated vectors; and then go further to estimate the VECM and do innovation accounting based on some cointegrating restrictions - in the style of Common Trends Approach. Bye Francis On 13 Feb 97 at 11:50, Dick Durevall Dick Durevall wrote: > From: Dick Durevall > To: "RATS Discussion List" > Subject: Re: cointegration: pi matrix of full rank > Date: Thu, 13 Feb 1997 11:50:18 +0100 > Reply-to: "RATS Discussion List" > If the pi matrix is of *full rank* then all variables are stationary, i.e., > you have five cointegratring vectors, and you can estimate the VAR in levels. > > Dick > > At 11.12 1997-02-12 EST, you wrote: > > > >This is more of a substantive question than a technical one, but here > >goes... > > > >Using CATS (johansen ml procedure), I have tested for cointegration in a > >5-variable system. > >According to the L-max and L-trace tests, the pi matrix is of *full rank* > >and my diagnostic tests > >indicate white noise residuals. > > > >It would appear that my next step to estimate a VAR in differences and > >then > >identify equations for innovation accounting? Can anyone confirm this > >for me? I can't > >find any examples in the literature of this happening (although I can > >understand why if > >the researcher is indeed attempting to discover a cointegrating > >relationship...) > > > >PS: Feel free to be as verbose as possible. This is my first attempt > >with the Johansen procedure > >and I could use all the advice I can get. (I already have Enders 1995) > > > >Thanks a million, > >Jonathan Deitch > >Pentagon Federal Credit Union > >Alexandria, VA USA > > > >email: b0deitch@hq.PenFed.org > > > > > > **~~**~~**~~**~~**~~**~~**~~**~~**~~****~~**~~**~~**~~** Francis Y. Kumah CentER for Economic Research P. O. Box 90153 5000 LE Tilburg The Netherlands Phone +31 13 466 2678 Fax +31 13 466 3066 My Home Page: http://cwis.kub.nl/~few5/center/phd_stud/kumah/home.HTM **~~**~~**~~**~~**~~**~~**~~**~~**~~****~~**~~**~~**~~** ---------- End of message ---------- From: "Estima" To: "RATS Discussion List" Subject: Re: discussion list archives on the Web Date: Thu, 13 Feb 1997 10:47:19 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v2.42) (via Mercury MTS (Bindery) v1.30) > Chris > > > The RATS-L discussion list archives may be viewed on the Web from the > > URL > > > > http://fmwww.bc.edu/ec-p/ratsL.html > > ... > > Well done and many thanks. Perhaps Tom can add a link from Estima's home page? > > Cheers > > Rob > Done! Links are available from a couple spots on our web page, and I spiffed up the page on the mailing list a bit. Thanks for all the work on the archives. The upcoming issue of our "RATSLetter" newsletter includes info on the mailing list. It should start hitting mailboxes around the 1st week of March, so hopefully we'll see more subscribers coming on board in early March. Sincerely, Tom Maycock Estima ---------- End of message ---------- From: oliver To: "RATS Discussion List" Subject: multivariate garch Date: Thu, 13 Feb 97 11:02:33 CST Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Mercury MTS (Bindery) v1.30 Hi: Is there anybody know whether the RATS can handle the Trivariate-Garch model? Thank you very much for your help. Oliver Rui Dept of Finance University of Houston 713-743-4783 oliver@rics1.cba.uh.edu ---------- End of message ---------- From: Gonzalo Camargo To: "RATS Discussion List" Subject: We need help Date: Thu, 13 Feb 1997 16:56:54 -0500 (GMT-0500) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Dear friends: We are Giovanna Aguilar and Gonzalo Camargo, we are researches at the Economic Department of the Catholic University of Peru. Now, we are preparing a paper that is based in a structural VAR, we are interested in modelling the relation between Lima (Peru"s capital city) and the rest of cities in the country. We would like to apply the Blanchard and Quah decomposition. We are using the econometric software named RATS 4.20 for windows, but this software does not have an especific program for make the Blanchard and Quah Variance Decomposition that we want to use. If you have used this methodology, could you send us the program used? We would like to receive your answer as soon as possible, please. Thank you for your attention. Gonzalo Camargo Guiovanna Aguilar ---------- End of message ---------- From: Giovanna Aguilar Andia To: "RATS Discussion List" Subject: About Blanchard & Quah Variance Decomposition Date: Thu, 13 Feb 1997 16:59:03 -0500 (GMT-0500) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Hi, Is there anybody know whether the RATS can make the Blanchard & Quah Variance decomposition? Thank you very much for your help, Giovanna Aguilar ---------- End of message ---------- From: Matysiak GA To: "RATS Discussion List" Subject: Re:Bootstrapping Correlations Date: Thu, 13 Feb 1997 22:23:34 +0000 (GMT) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 As part of a bootstrapping procedure I wish to undertake the following task: 1) Select,say, 3 time series at random from 30 series, with replacement, and create a new series which is the average of the 3, being the series seravg 2) Calculate the correlation coefficient between seravg and seravg lagged by one period 3) Store the result 4) Repeat the above, say, 1000 times I will then look at the distribution of the correlation coeffs. Any suggestions as to how I could select two equal random subsets from the thirty and obtain the correlations would be appreciated. Many thanks. George Matysiak City University Business School London ---------- End of message ---------- From: "Christopher F. Baum" To: "RATS Discussion List" Subject: bootstrapping correlations Date: Thu, 13 Feb 1997 17:44:25 -0400 (EDT) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 * in reply to G.A. Matysiak enquiry posted to RATS-L * here's the first part, generating a random set of three series * to be analysed from the set of 30 * comp nser=30 comp nrep=10 comp ntot=nser+nrep all ntot 20 do i=1,nser set [series]i = i**2 enddo dec rect[integer] s(nrep,3) dec rect serno(nrep,3) mat serno=%uniform(1,nser) ewise s(i,j)=fix(serno(i,j)+0.49) write s do i=1,nrep set [series](i+nser) = ([series]s(i,1)+[series]s(i,2)+[series]s(i,3))/3.0 enddo tab end Kit Baum baum@bc.edu ---------- End of message ---------- From: David Aadland To: "RATS Discussion List" Subject: Re: We need help Date: Thu, 13 Feb 1997 15:17:43 -0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Windows Eudora Light Version 1.5.2 (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Giovanna and Gonzala: You are in luck! Go to Estima's homepage (www.estima.com) and download the SVAR and VMA procedures. I've had success using these procedures to estimate Blanchard and Quah's structural VAR and impulse response functions. You will find Carlo Giannini' David Aadland University of Oregon At 04:56 PM 2/13/97 -0500, you wrote: >Dear friends: > >We are Giovanna Aguilar and Gonzalo Camargo, we are researches >at the Economic Department of the Catholic University of Peru. > >Now, we are preparing a paper that is based in a structural VAR, we are interested in modelling the relation between Lima (Peru"s >capital city) and the rest of cities in the country. We would like to >apply the Blanchard and Quah decomposition. > >We are using the econometric software named RATS 4.20 for windows, but >this software does not have an especific program for make the Blanchard >and Quah Variance Decomposition that we want to use. >If you have used this methodology, could you send us the program used? > >We would like to receive your answer as soon as possible, please. > >Thank you for your attention. > >Gonzalo Camargo Guiovanna Aguilar > > > > ---------- End of message ---------- From: WATT WING HONG To: "RATS Discussion List" Subject: RE: multivariate garch Date: Fri, 14 Feb 1997 07:55:14 +0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63 (via Mercury MTS (Bindery) v1.30) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit RATS 4.20 allows one to specifiy a multivariate-GARCH. You can implement either the vech version, the BEKK version or the constant correlation version. -- W.H.Watt Division of Banking & Finance Nanyang Business School Nanyang Technological University Singapore >Hi: >Is there anybody know whether the RATS can handle the Trivariate-Garch model? >Thank you very much for your help. > >Oliver Rui >Dept of Finance >University of Houston >713-743-4783 >oliver@rics1.cba.uh.edu > ---------- End of message ---------- From: Giovanna Aguilar Andia To: "RATS Discussion List" Subject: About procedures Date: Thu, 13 Feb 1997 20:18:20 -0500 (GMT-0500) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 ---------- Forwarded message ---------- Date: Mon, 10 Feb 1997 15:24:20 -0500 (GMT-0500) From: Giovanna Aguilar Andia To: Subject: About procedures (fwd) Dear Friends: My name is Giovanna Aguilar, and I am working in the Department of Economics at the "Universidad Catolica del Peru" as profesor and reasercher. Actually I am working in a reaserch about cycles and trends in peruvian economy. I am trying to find the cycle and trend of a group of macroeconomic series that have a common trend (they are cointegrated) and perhaps a common cycle. To perform this task, I need to use the Vahid and Engle methodology. I would like to know if you have a procedure with this methodology. In that case I would apreciate your sending me the information of where and how I can obtain it. The paper of Vahid and Engle is: F. Vahid and R. F. Engle. "Common trends and common cycles". In: Journal of Applied Econometrics. Vol.8, 341-360 (1993) I will be very grateful for your help. Sincerely Giovanna Aguilar ---------- End of message ---------- From: David Aadland To: "RATS Discussion List" Subject: Re: We need help Date: Thu, 13 Feb 1997 17:43:23 -0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Windows Eudora Light Version 1.5.2 (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Giovanna and Gonzala: You are in luck! Go to Estima's homepage (www.estima.com) and download the SVAR and VMA procedures. I've had success using these procedures to estimate Blanchard and Quah's structural VAR and impulse response functions. You will find Carlo Giannini' David Aadland University of Oregon At 04:56 PM 2/13/97 -0500, you wrote: >Dear friends: > >We are Giovanna Aguilar and Gonzalo Camargo, we are researches >at the Economic Department of the Catholic University of Peru. > >Now, we are preparing a paper that is based in a structural VAR, we are interested in modelling the relation between Lima (Peru"s >capital city) and the rest of cities in the country. We would like to >apply the Blanchard and Quah decomposition. > >We are using the econometric software named RATS 4.20 for windows, but >this software does not have an especific program for make the Blanchard >and Quah Variance Decomposition that we want to use. >If you have used this methodology, could you send us the program used? > >We would like to receive your answer as soon as possible, please. > >Thank you for your attention. > >Gonzalo Camargo Guiovanna Aguilar > > > > ---------- End of message ---------- From: wan@hpap.tku.edu.tw To: "RATS Discussion List" Subject: fractional integrated! Date: Fri, 14 Feb 1997 12:46:05 +0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Mozilla 3.01 (Win95; I) (via Mercury MTS (Bindery) v1.30) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit is there anyone who has the procedure to estimate FIGARCH model under RATS? thanks ---------- End of message ---------- From: "Ward, Bert D" To: "RATS Discussion List" Subject: QMLE Date: Mon, 17 Feb 1997 16:32:58 +1300 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: Lincoln University X-Mailer: Pegasus Mail v3.31 (via Mercury MTS (Bindery) v1.30) Content-transfer-encoding: 7BIT Does anybody have any instructions on how to code for Quasi-Maximum Likelihood Estimation for Multivariate GARCH models in RATS (refer Hamilton (1994) pp.663- 664)? Kind regards Bert D. Ward Senior Lecturer in Economics Dept. of Economics & Marketing PO Box 84 Lincoln University Canterbury NEW ZEALAND Telephone (64)(3) 325 2811 extn 8261 FAX (64)(3) 325 3847 ---------- End of message ---------- From: "vmurthy" To: "RATS Discussion List" Subject: Hypotheses Testing Using Cats Date: Mon, 17 Feb 1997 08:33:56 CST Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Organization: Creighton University X-Mailer: Pegasus Mail/Windows (v1.22) (via Mercury MTS (Bindery) v1.30) Does any one have a sample problem that illustrates hypotheses testing ,with one or multiple cointegrating vectors, using CATS? If so, please send me a copy. Are there any RATS procedures that can be used to implement the Phillips-Hansen FM estimation technique and the Hansen's c.v.test? If so, please let me know the details. Thanks in advance. V.Murthy ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: Re: QMLE Date: Tue, 18 Feb 1997 10:10:22 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 Bert > Does anybody have any instructions on how to code for Quasi-Maximum > Likelihood Estimation for Multivariate GARCH models in RATS (refer > Hamilton (1994) pp.663- 664)? I'm assuming that your question is really about QMLE GARCH rather than Multivariate GARCH per se. My apologies if I've got that premise wrong! If you look at this carefully, I think that you will find that it essentially involves the matrix of first derivatives (outer-product of the gradient vector) and the matrix of second derivatives (the Hessian matrix). Weiss, Bollerslev and Woolridge, Lumsdain, et al all provide analytical recursive formulas for computing the various terms for particular xARCH specifications. While my memory suggests that these are all univariate specifications, I suspect that with a sharp pencil and some paper one should be able to derive the corresponding formulas for the multivariate specifications of interest to you. You could then code these up in RATS - and post your code to this list. Personally, I'm content to do these calculations numerically. If you look at the RATS manual I think you'll find that something like: MAXIMISE(METHOD=BFGS,ROBUSTERRORS,RECURSIVE,ITERATIONS=100) LOGL start end will provide you with what you want for any specification - provided, of course, that the BFGS algorithm does 'enough' iterations to converge to the Hessian. (Just don't ask me what is 'enough'!) [For what its worth, I can say that the Fortran code used by McCurdy and Morgan for a number of papers also computes the QMLE covariance matrix numerically rather than analytically - or at least the version used for those papers did.] Hope this helps Rob ----------------------------------------------------------------------- Rob Trevor Associate Professor Centre for Studies in Money, Banking and Finance Macquarie University, NSW, AUSTRALIA 2109 Internet: robt@efs.mq.edu.au VOICE: +61 (2) 9850-8447 FAX: +61 (2) 9850-7281 or +61 (2) 9418-1847 ----------------------------------------------------------------------- ---------- End of message ---------- From: WATT WING HONG To: "RATS Discussion List" Subject: RE: QMLE Date: Tue, 18 Feb 1997 07:58:03 +0800 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63 (via Mercury MTS (Bindery) v1.30) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit >Personally, I'm content to do these calculations numerically. If you look >at the RATS manual I think you'll find that something like: > >MAXIMISE(METHOD=BFGS,ROBUSTERRORS,RECURSIVE,ITERATIONS=100) LOGL start end > >will provide you with what you want for any specification - provided, of >course, that the BFGS algorithm does 'enough' iterations to converge to the >Hessian. (Just don't ask me what is 'enough'!) If I recall correctly, 'enough' means doing as many iterations as there are parameters in the system. Once, I made the mistake of doing the BHHH estimation before the BFGS estimation, and the standard errors of the latter didn't make sense. The 'trick' is to do the BFGS estimation before the BHHH estimation - that way, one can get both robust t-statistics as well as asymptotic ones. -- W H Watt Division of Banking & Finance Nanyang Business School Nanyang Technological University Singapore Email: awhwatt@ntu.edu.sg > ---------- End of message ---------- From: Rob Trevor To: "RATS Discussion List" Subject: RE: QMLE Date: Tue, 18 Feb 1997 11:10:30 +1100 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 At 10:58 AM +1100 18/2/97, WATT WING HONG wrote: > >Personally, I'm content to do these calculations numerically. If you look > >at the RATS manual I think you'll find that something like: > > > >MAXIMISE(METHOD=BFGS,ROBUSTERRORS,RECURSIVE,ITERATIONS=100) LOGL start end > > > >will provide you with what you want for any specification - provided, of > >course, that the BFGS algorithm does 'enough' iterations to converge to the > >Hessian. (Just don't ask me what is 'enough'!) > > If I recall correctly, 'enough' means doing as many iterations > as there are parameters in the system. Once, I made the mistake ....only if the objective function is a quadratic. I'm not aware of any results for the kind of log-likelihood we have here, although as the manual says, you'd probably want to do at least as many as this... > of doing the BHHH estimation before the BFGS estimation, and > the standard errors of the latter didn't make sense. The 'trick' > is to do the BFGS estimation before the BHHH estimation - that > way, one can get both robust t-statistics as well as asymptotic > ones. Precisely. [As a side 'benefit' you also get to see how doing one more iteration changes your parameter estimates.] Cheers Rob ---------- End of message ---------- From: "L.Saldanha" To: "RATS Discussion List" Subject: RE: QMLE Date: Thu, 20 Feb 1997 12:52:07 +0000 (GMT) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 Hi, I'd like to test for the equality of the coefficients from a linear regression and another to test if the autoregressive coefficients sum to unity. Would be grateful for any suggestions, 'summarise' and 'test' do not quite work for these tests. Thank you Saldanha ---------- End of message ---------- From: "ABDOL S. SOOFI, ECONOMICS, UW-PLATTEVILLE; HTTP://VMS.WWW.UWPLATT.EDU/~SOOFI" To: "RATS Discussion List" Subject: Maximum-Likelihood Estimation of Multimodal Functions Date: Sat, 22 Feb 1997 13:33:53 -0600 (CST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 > > >Personally, I'm content to do these calculations numerically. If you look > > >at the RATS manual I think you'll find that something like: > > > > > >MAXIMISE(METHOD=BFGS,ROBUSTERRORS,RECURSIVE,ITERATIONS=100) LOGL start end > > > > > >will provide you with what you want for any specification - provided, of > > >course, that the BFGS algorithm does 'enough' iterations to converge to the > > >Hessian. (Just don't ask me what is 'enough'!) > > > > If I recall correctly, 'enough' means doing as many iterations > > as there are parameters in the system. Once, I made the mistake > > ...only if the objective function is a quadratic. I'm not aware of any > results for the kind of log-likelihood we have here, although as the manual > says, you'd probably want to do at least as many as this... > > > of doing the BHHH estimation before the BFGS estimation, and > > the standard errors of the latter didn't make sense. The 'trick' > > is to do the BFGS estimation before the BHHH estimation - that > > way, one can get both robust t-statistics as well as asymptotic > > ones. > > Precisely. [As a side 'benefit' you also get to see how doing one more > iteration changes your parameter estimates.] > I was wondering whether the above procedures (algorithms) can deal with multimodal functions. Keep in mind that maximizing the log-likelihood functions for many economic and financial time series invariably involves multimodality. Please visit my home page for more information on this issue, if you find my question at relevant. Regards, --Abdol *************************************************************************** *************************************************************************** Abdol S. SOOFI Department of Economics University of Wisconsin-Platteville Platteville, WI 53818-3099 USA (608) 342-1570 Soofi@uwplatt.edu URL: http://vms.www.uwplatt.edu/~soofi ---------- End of message ---------- From: "ABDOL S. SOOFI, ECONOMICS, UW-PLATTEVILLE; HTTP://VMS.WWW.UWPLATT.EDU/~SOOFI" To: "RATS Discussion List" Subject: Maximum-likelihood Function Estimation and Multimodality Date: Sat, 22 Feb 1997 13:50:27 -0600 (CST) Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII X-Mailer: Mercury MTS (Bindery) v1.30 > From: IN%"RATS-L@efs.mq.edu.au" "RATS Discussion List" 22-FEB-1997 13:35:49.70 > To: IN%"RATS-L@efs.mq.edu.au" "RATS Discussion List" > CC: > Subj: Maximum-Likelihood Estimation of Multimodal Functions > > Return-path: > Received: from ness.efs.mq.edu.au by uwplatt.edu (PMDF V5.1-5 #17356) > with ESMTP id <01IFPVF2ICGWBL6BHG@uwplatt.edu> for SOOFI; Sat, > 22 Feb 1997 13:35:41 CST > Received: from efs1.efs.mq.edu.au (efs1.efs.mq.edu.au [137.111.64.8]) > by ness.efs.mq.edu.au (8.8.4/8.8.3) with ESMTP id GAA08118; Sun, > 23 Feb 1997 06:32:04 +1100 (EST) > Received: from EFS1/SpoolDir by efs1.efs.mq.edu.au (Mercury 1.21); Sun, > 23 Feb 1997 06:32:01 +0000 > Received: from SpoolDir by EFS1 (Mercury 1.30); Sun, 23 Feb 1997 06:31:42 +0000 > Date: Sat, 22 Feb 1997 13:33:53 -0600 (CST) > From: > "ABDOL S. SOOFI, ECONOMICS, UW-PLATTEVILLE; HTTP://VMS.WWW.UWPLATT.EDU/~SOOFI" > > Subject: Maximum-Likelihood Estimation of Multimodal Functions > Sender: Maiser@efs1.efs.mq.edu.au > To: RATS Discussion List > Errors-to: rats-l-owner@efs.mq.edu.au > Reply-to: RATS Discussion List > Message-id: <1D5B99745DA@efs1.efs.mq.edu.au> > MIME-version: 1.0 > X-Mailer: Mercury MTS (Bindery) v1.30 > Content-type: TEXT/PLAIN; CHARSET=US-ASCII > X-Listname: > > > > > >Personally, I'm content to do these calculations numerically. If you look > > > >at the RATS manual I think you'll find that something like: > > > > > > > >MAXIMISE(METHOD=BFGS,ROBUSTERRORS,RECURSIVE,ITERATIONS=100) LOGL start end > > > > > > > >will provide you with what you want for any specification - provided, of > > > >course, that the BFGS algorithm does 'enough' iterations to converge to the > > > >Hessian. (Just don't ask me what is 'enough'!) > > > > > > If I recall correctly, 'enough' means doing as many iterations > > > as there are parameters in the system. Once, I made the mistake > > > > ...only if the objective function is a quadratic. I'm not aware of any > > results for the kind of log-likelihood we have here, although as the manual > > says, you'd probably want to do at least as many as this... > > > > > of doing the BHHH estimation before the BFGS estimation, and > > > the standard errors of the latter didn't make sense. The 'trick' > > > is to do the BFGS estimation before the BHHH estimation - that > > > way, one can get both robust t-statistics as well as asymptotic > > > ones. > > > > Precisely. [As a side 'benefit' you also get to see how doing one more > > iteration changes your parameter estimates.] > > > > I was wondering whether the above procedures (algorithms) can deal with multimodal > functions. Keep in mind that maximizing the log-likelihood functions for many > economic and financial time series invariably involves multimodality. > > Please visit my home page for more information on this issue, if you find my > question at relevant. > > Regards, > > --Abdol > > > *************************************************************************** > *************************************************************************** > > Abdol S. SOOFI > Department of Economics > University of Wisconsin-Platteville > Platteville, WI 53818-3099 USA > > (608) 342-1570 > Soofi@uwplatt.edu > URL: http://vms.www.uwplatt.edu/~soofi *************************************************************************** *************************************************************************** Abdol S. SOOFI Department of Economics University of Wisconsin-Platteville Platteville, WI 53818-3099 USA (608) 342-1570 Soofi@uwplatt.edu URL: http://vms.www.uwplatt.edu/~soofi ---------- End of message ---------- From: Eric Bentzen To: "RATS Discussion List" Subject: Date: Mon, 24 Feb 97 8:28:45 MET Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Mercury MTS (Bindery) v1.30 help ---------- End of message ---------- From: Economic Group To: "RATS Discussion List" Subject: Mac Rats Date: Wed, 26 Feb 1997 10:44:18 +1000 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Mailer: Mercury MTS (Bindery) v1.30 Tom, Do we have an updated release date or is it still vapourware?? Mark :-{) Economic Analysis Department Email enquiries: Economic Group PH: +612 9551 8817 Reserve Bank of Australia FAX: +612 9551 8033 ---------- End of message ---------- From: "Estima" To: "RATS Discussion List" Subject: Re: Mac Rats Date: Wed, 26 Feb 1997 10:03:02 -0600 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT X-mailer: Pegasus Mail for Win32 (v2.42) (via Mercury MTS (Bindery) v1.30) > > Do we have an updated release date or is it still vapourware?? > Well, MacRATS isn't vapourware (i.e. it exists, it just ain't done yet), but no, I don't have a release date yet. I'll be sure to post to the list when I do. Sincerely, Tom Maycock Estima -- +-----------------------------+-----------------------------------------+ | Estima | | | P.O. Box 1818 | Voice: (847) 864-8772 | | Evanston, IL 60204-1818 | Fax: (847) 864-6221 | | U.S.A | BBS: (847) 864-8816 | | e-mail: estima@estima.com | CompuServe: 73140,2202 | |-----------------------------------------------------------------------| | Web Site: http://www.estima.com | | RATS Internet Mailing List: New members can join by sending e-mail to | | MAISER@EFS.MQ.EDU.AU with the message: SUBSCRIBE RATS-L | +-----------------------------------------------------------------------+ ---------- End of message ----------