From: Wai Lee To: "RATS Discussion List" Subject: -No Subject- Date: 2 Dec 96 17:17:51 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: Mime-Version: 1.0 Content-Type: Text/Plain X-Mailer: Mercury MTS (Bindery) v1.30 I have a daily xls file with dates as first column, with holidays and weekends skipped. I do CALENDAR 1980 1 2 ALL 1996:9:30 OPEN DATA ÉÉÉÉ. DATA(FORMAT=XLS,ORG=OBS,VERBOSE) / ÉÉÉÉÉ. etc. I suppose RATS will fill in "NA" for me for those skipped dates, as described on p2-20 of manual. But it doesn't. Instead, it correctly reads the data in Jan of 80, but all Feb of 80 data become "NA". Data for 1980:2:1 is read as the data for 1980:3:1 and the like. It simply mess up the whole data file. Have I done anything wrong? When I use VERBOSE option of DATA, it shows that RATS read number of rows and columns correctly, but it also indicates "Compact Daily(7) from 1980:1:2 to Daily(5) 1980:1:2" while I do not use COMPACT option at all. Thanks for any help. P.S. Any more information about web application of Unix RATS? ---------- End of message ---------- From: "Estima" To: "RATS Discussion List" Subject: Re: -No Subject- Date: Tue, 3 Dec 1996 10:47:11 -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 Windows (v2.31) (via Mercury MTS (Bindery) v1.30) > I have a daily xls file with dates as first column, with holidays and weekends > skipped. I do > > CALENDAR 1980 1 2 > ALL 1996:9:30 > OPEN DATA ÉÉÉÉ. > DATA(FORMAT=XLS,ORG=OBS,VERBOSE) / ÉÉÉÉÉ. etc. > > I suppose RATS will fill in "NA" for me for those skipped dates, as described > on p2-20 of manual. But it doesn't. Instead, it correctly reads the data in > Jan of 80, but all Feb of 80 data become "NA". Data for 1980:2:1 is read as > the data for 1980:3:1 and the like. It simply mess up the whole data file. > Dear Wai: First, I assume your CALENDAR instruction was actually: CALENDAR(DAILY) 1980 1 2 > > When I use VERBOSE option of DATA, it shows that RATS read number of rows and > columns correctly, but it also indicates > > "Compact Daily(7) from 1980:1:2 to Daily(5) 1980:1:2" > > while I do not use COMPACT option at all. > When reading daily data from a spreadsheet, RATS treats the file as having seven-day data. In other words, it is allowing for the possibility that the spreadsheet file may contain some weekend data. If the weekend dates are not present on the file, that's fine--the series in memory will simply be padded with NA's if you are using CAL with the SEVENDAY option, or omitted completely if you are using the DAILY (i.e. 5-day) option. The message from VERBOSE is just telling you that the source data is daily and that the CALENDAR is set for 5-day per week daily. No "compaction" is being done, other than the fact that weekend data, if present, is being ignored due to the current CALENDAR setting. I did some quick testing, and couldn't reproduce any problems reading data from XLS files (i.e. observations such as holidays that were not present on the XLS file were properly set to NA in RATS). If you can't resolve the problem, perhaps you could send me a copy of the data file and program so I can try them out here. 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: Wai Lee To: "RATS Discussion List" Subject: geometric average Date: 5 Dec 96 16:24:40 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 X-Mailer: Mercury MTS (Bindery) v1.30 I may have been careless, but I cannot find an easy way to compute the geometric average of a series, instead of the arithmetic average from the RATS manual. Is there such a command? I hope I don't need to use a do loop (that's what I am using). And let's say, I don't want to take ln. Thanks for suggestion. ---------- End of message ---------- From: Sean Collins To: "RATS Discussion List" Subject: geometric average -Reply Date: Fri, 06 Dec 1996 10:46:27 +1200 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) Take the arithmetic average of the logs and re-exponentiate? ---------- End of message ---------- From: David Aadland To: "RATS Discussion List" Subject: Re: geometric average Date: Thu, 05 Dec 1996 19:24:36 -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" Use COMPACT=GEOMETRIC as an option on the DATA command. David Aadland University of Oregon At 04:24 PM 12/5/96 EDT, you wrote: >I may have been careless, but I cannot find an easy way to compute the >geometric average of a series, instead of the arithmetic average from the RATS >manual. Is there such a command? I hope I don't need to use a do loop (that's >what I am using). And let's say, I don't want to take ln. > >Thanks for suggestion. > > > ---------- End of message ---------- From: Wai Lee To: "RATS Discussion List" Subject: geometric average Date: 6 Dec 96 8:49:50 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 X-Mailer: Mercury MTS (Bindery) v1.30 Thanks for suggestions from Aadland and Collins. I think both methods will work. I don't mean to be troublesome, but I was just wondering whether I have missed a simple command in RATS that can give me the geometric average. The COMPACT = GEOMETRIC may not be appropriate for me because the time series is internally generated from my program, rather than from an external file. May be I will just use ln, if it is the best choice. Thanks. ---------- End of message ---------- From: RAY DEGENNARO To: "RATS Discussion List" Subject: p-values Date: Mon, 16 Dec 96 22:52:15 LCL Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Mercury MTS (Bindery) v1.30 How can I output the significance levels (p-values) of the coefficients from MAXIMIZE? I get the t-statistics, of course, but I want the p-values for input into another program. The %SIGNIF function only seems to work for a single test. Furthermore, I have no need to use any of the 5 required test instructions and the program is plenty long already. I'm hoping that there is a function that returns the p-value of the coefficient vector. For example, if I have 3 coefficients, it would be great to just reference them as %wishlist(1) %wishlist(2) %wishlist(3) Does anyone know how to do this, or some other way to retrieve p-values? If possible, please reply to rdegenna@utk.edu as well as the discussion list, as I sometimes don't get messages from the list. Thanks. ***************************************************************** Ramon P. DeGennaro Department of Finance 434 Stokely Management Center The University of Tennessee Knoxville, TN 37996-0540 e-mail rdegenna@utk.edu ***************************************************************** ---------- End of message ---------- From: jacques tebeka <101677.76@compuserve.com> To: "RATS Discussion List" Subject: estimation of STAR model Date: Tue, 24 Dec 1996 12:45:26 -0500 Errors-to: Reply-to: "RATS Discussion List" Sender: Maiser@efs1.efs.mq.edu.au X-listname: X-Mailer: Mercury MTS (Bindery) v1.30 Is there anyone who has a solution to estimate LSTAR and ESTAR model (Granger & Terasvirta, 1993). I tried to use the instruction NLLS but i havE always NA for the parameters or no convergence 5THEN I USED NLPAR but without success). May be someone had the same problem, if you can help meÉ Thanks. TEBEKA JACQUES GROUP ESSEC, FRANCE e-mail: 101677.76@compuserve.com ---------- End of message ----------