Return-Path: <maiser@efs.mq.edu.au>
Received: from jules.bc.edu (jules.bc.edu [136.167.2.173])
	by monet.bc.edu (8.8.7/8.8.7) with ESMTP id KAA86070
	for <BAUM@monet.bc.edu>; Sat, 15 Nov 1997 10:42:41 -0500
From: maiser@efs.mq.edu.au
Received: (from root@localhost)
	by jules.bc.edu (8.8.7/8.8.7) with X.500 id KAA103340
	for BAUM@MAIL1.BC.EDU; Sat, 15 Nov 1997 10:42:40 -0500
Received: from baldrick.ocs.mq.edu.au (baldrick.ocs.mq.edu.au [137.111.1.12])
	by jules.bc.edu (8.8.7/8.8.7) with ESMTP id KAA66211
	for <baum@bc.edu>; Sat, 15 Nov 1997 10:42:34 -0500
Received: from efs1.efs.mq.edu.au (efs1.efs.mq.edu.au [137.111.64.8])
	by baldrick.ocs.mq.edu.au (8.8.5/8.8.5) with ESMTP id CAA09715
	for <baum@bc.edu>; Sun, 16 Nov 1997 02:42:30 +1100 (EST)
Received: from EFS1/SpoolDir by efs1.efs.mq.edu.au (Mercury 1.21);
    16 Nov 97 02:41:35 GMT+1000
Received: from SpoolDir by EFS1 (Mercury 1.30); 16 Nov 97 02:41:34 GMT+1000
To: baum@bc.edu
Date: Sun, 16 Nov 97 2:41:34 GMT+1000
Subject: Re: 
Message-ID: <40E4300999@efs1.efs.mq.edu.au>

From: Rodolfo =?iso-8859-1?Q?Berm=FAdez?= Neubauer <rodolfo@samerica.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Error with 600 Observations
Date: Mon, 03 Nov 1997 17:47:31 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: QUALCOMM Windows Eudora Light Version 3.0.2 (32) (via Mercury MTS (Bindery) v1.30)
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

I'm trying to multiply a matrix of 1x599 with one of 599x5999 and as soon
as i run the program WinRATS will "think" for a while and then shutdown
with an application error. We're using WinRATS 4.2 on MS Windows 95. If
anyone knows of this bug and a way of coorecting it, please notify me as
soon as possible.

Thank you in advance for your valueble collaboration.

Rodolfo BErmudez Neubauer


---------- End of message ----------

From: Gongpil Choi <gpchoi@sun.kif.re.kr>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: UNSUBSCRIBE
Date: Wed, 05 Nov 1997 09:50:37 +0900
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: KIF
X-Mailer: Mozilla 3.01Gold (Win95; I) (via Mercury MTS (Bindery) v1.30)
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Please delete my name from your mailing list. Thanks.

---------- End of message ----------

From: Byung Chul Ahn <ahnbc@ynucc.yeungnam.ac.kr>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Job Openings at Yeungnam University
Date: Wed, 05 Nov 1997 11:18:15 +0900
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Yeungnam University
X-Mailer: Mozilla 4.01 [en] (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 8bit

Dear Rats users;
Please excuse this announcement which is not directly related with 
this mailing list.  We have openings 4 professorships as following:

Yeungnam University, Republic of Korea

H0 Public finance
J0 Labor economics
E0 Monetary economics
F0 International Finance

  The Department of Economics invites applications for 4 professorships
for 1998 academic year starting March 1, 1998.  The openings are for 
non-Korean nationals.  The contract is for 2 years
and renewable after initial appointment.  The Department requires
minimal teaching load of 9 hours per week each semester.  We wish to
hire in the fields of, but not limited to, labor economics, monetary
economics, public finance and international finance.  Applicants should
be able to teach microeconomics or macroeconomics to the undergraduate
students in English.  The Ph.D. is required at time of appointment.
  Benefits: Appointment carries a gross stipend of between US$30,000 and
US$35,000 depending exchange rates.  University subsidized apartment is
available at nominal rent (about US$50).  The department will provide
Pentium based PC and laser printer for new appointees.  Other assistance
would be available if needed.
  To apply, submit curriculum vita, 2 letters of recommendation and
copies of published and/or unpublished work to the contact no later than
December 10, 1997.  CONTACT: G. L. Son, Director, Dept. of Economics,
Yeungnam University, 214-1 Dae-dong Kyungsan City, Kyungbuk 712-749
Korea. (email:glson@ynucc.yeungnam.ac.kr, Phone:+82-53-810-2711,
Fax:+82-53-812-5321)


-- 
Byung Chul Ahn, Department of Economics, Yeungnam University
214-1 Dae-dong, Kyungsan, Kyungbuk 712-749 Korea
Phone : +82-53-810-2726            Fax : +82-53-812-5321
mailto:ahnbc@ynucc.yeungnam.ac.kr  http://ynucc.yeungnam.ac.kr/~ahnbc/

---------- End of message ----------

From: Byung Chul Ahn <ahnbc@ynucc.yeungnam.ac.kr>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Job Openings at Yeungnam University
Date: Wed, 05 Nov 1997 13:59:52 +0900
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Yeungnam University
X-Mailer: Mozilla 4.01 [en] (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 8bit

Dear Rats users;

We forgot to mention department home pages for further information.
Sorry for causing inconveniences.  Address added at the end of this
announcement.

Job Openings at Yeungnam University, Republic of Korea

H0 Public finance
J0 Labor economics
E0 Monetary economics
F0 International Finance

  The Department of Economics invites applications for 4 professorships
for 1998 academic year starting March 1, 1998.  The openings are for 
non-Korean nationals.  The contract is for 2 years
and renewable after initial appointment.  The Department requires
minimal teaching load of 9 hours per week each semester.  We wish to
hire in the fields of, but not limited to, labor economics, monetary
economics, public finance and international finance.  Applicants should
be able to teach microeconomics or macroeconomics to the undergraduate
students in English.  The Ph.D. is required at time of appointment.
  Benefits: Appointment carries a gross stipend of between US$30,000 and
US$35,000 depending exchange rates.  University subsidized apartment is
available at nominal rent (about US$50).  The department will provide
Pentium based PC and laser printer for new appointees.  Other assistance
would be available if needed.
  To apply, submit curriculum vita, 2 letters of recommendation and
copies of published and/or unpublished work to the contact no later than
December 10, 1997.  CONTACT: G. L. Son, Director, Dept. of Economics,
Yeungnam University, 214-1 Dae-dong Kyungsan City, Kyungbuk 712-749
Korea. (mailto:glson@ynucc.yeungnam.ac.kr, Phone:+82-53-810-2711,
Fax:+82-53-812-5321)

For further information, you may visit department home page at
http://ynucc.yeungnam.ac.kr/~econ/jobopening.html

-- 
Byung Chul Ahn, Department of Economics, Yeungnam University
214-1 Dae-dong, Kyungsan, Kyungbuk 712-749 Korea
Phone : +82-53-810-2726            Fax : +82-53-812-5321
mailto:ahnbc@ynucc.yeungnam.ac.kr  http://ynucc.yeungnam.ac.kr/~ahnbc/

---------- End of message ----------

From: DAVID BYERS <ByersJD@cardiff.ac.uk>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Error with 600 Observations
Date: Thu, 6 Nov 1997 11:56:30 GMT0BST
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-mailer: Pegasus Mail for Windows (v2.54) (via Mercury MTS (Bindery) v1.30)

> From:          Rodolfo Bermudez Neubauer <rodolfo@samerica.com>
> To:            RATS Discussion List <RATS-L@efs.mq.edu.au>
> Subject:       Error with 600 Observations
> Date:          Mon, 03 Nov 1997 17:47:31 -0500
> Reply-to:      RATS Discussion List <RATS-L@efs.mq.edu.au>

> I'm trying to multiply a matrix of 1x599 with one of 599x5999 and as soon
> as i run the program WinRATS will "think" for a while and then shutdown
> with an application error. We're using WinRATS 4.2 on MS Windows 95. If
> anyone knows of this bug and a way of coorecting it, please notify me as
> soon as possible.
> 
I've had a similar outcome when estimating some nonlinear models. 
The programs run fine under RATS386. Slightly inconvenient, but worth 
a try.

 
David Byers
Cardiff Business School
Colum Drive
Cardiff CF1 3EU
UK
Tel + 1222 874526
Fax + 1222 874419

---------- End of message ----------

From: Rob Trevor <robt@efs.mq.edu.au>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: [ADMIN] Re: Job Openings at Yeungnam University
Date: Fri, 7 Nov 1997 02:55:46 +1100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
X-Mailer: Mercury MTS (Bindery) v1.30

Byung Chul Ahn wrote:
> Dear Rats users;
> Please excuse this announcement which is not directly related with
> this mailing list.  We have openings 4 professorships as following:
> ...

PLEASE do not do this...when you join this list you get an email explaining
the purpose of the list and reminding you that we're all busy people and do
not wish to waste our time dealing with extraneous email. If I allow people
to post such messages to this list, the list will die as many of the more
experienced RATS users will unsubscribe.

Please keep this list for RATS issues alone.

Rob Trevor
RATS-L Owner



---------- End of message ----------

From: Arturo =?iso-8859-1?Q?Jos=E9_?= Galindo Andrade <agalindo@students.uiuc.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Unbalanced Panels
Date: Thu, 06 Nov 1997 10:23:18 -0600
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Windows Eudora Pro Version 2.2 (32) (via Mercury MTS (Bindery) v1.30)
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

Hello. 
Could someone please suggest me a way of estimating an unbalanced panel -
random effects model using RATS?

In advance thanks,

Arturo Galindo
University of Illinois at Urbana-Champaign


---------- End of message ----------

From: "Ellis W. Tallman" <etallman@ipacific.net.au>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: [ADMIN] Re: Job Openings at Yeungnam University
Date: Fri, 07 Nov 1997 07:28:35 -0800
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.01 [en] (Win16; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thanks Rob.

Rob Trevor wrote:

> Byung Chul Ahn wrote:
> > Dear Rats users;
> > Please excuse this announcement which is not directly related with
> > this mailing list.  We have openings 4 professorships as following:
> > ...
>
> PLEASE do not do this...when you join this list you get an email
> explaining
> the purpose of the list and reminding you that we're all busy people
> and do
> not wish to waste our time dealing with extraneous email. If I allow
> people
> to post such messages to this list, the list will die as many of the
> more
> experienced RATS users will unsubscribe.
>
> Please keep this list for RATS issues alone.
>
> Rob Trevor
> RATS-L Owner




---------- End of message ----------

From: Felix.Maag@SBF.unisg.ch
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Covariance matrix
Date: Fri, 07 Nov 1997 14:31:12 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: text/plain; charset=us-ascii
Content-transfer-encoding: QUOTED-PRINTABLE
X-Mailer: Mercury MTS (Bindery) v1.30

Hi everyone.

I want to compute the kovariance matrix of severel time series.
Unfortunately rats doesn't compute what I expect.
Below you see the output of the covariance/correlation matrix. The di=
agonal
elements (variances) are completely
different from the variances I get for the individual time series. Fu=
rther
is the covariance and correlation different from what I get with
Eviews and Excel although I use the same data series.

Questions:
1) What's wrong?
2) How can I get a complete kovariance matrix ( without correlations)=
?

Looking forward for help. Thanks.

Felix


Covariance\Correlation Matrix
               RDI20                     RDI50
RDI20     0.000139920543      0.9661474545
RDI50     0.000131037751       0.000131469361

Statistics on Series RDI50
Monthly Data From 1988:02 To 1996:12
Observations   107
Sample Mean    0.00600369948              Variance          9.632519e=
-005
Standard Error 0.00981453968                  SE of Sample Mean
0.000949
t-Statistic        6.32763                         Signif Level (Mean=
=3D0)
0.00000001
Skewness          -0.20128                     Signif Level (Sk=3D0)
0.40198231
Kurtosis          -0.06765                         Signif Level (Ku=
=3D0)
0.89007747

Statistics on Series RDI20
Monthly Data From 1988:02 To 1996:12
Observations   107
Sample Mean    0.00603956290              Variance               0.00=
0104
Standard Error 0.01021861593             SE of Sample Mean      0.000=
988
t-Statistic        6.11372                         Signif Level (Mean=
=3D0)
0.00000002
Skewness          -0.57868                         Signif Level (Sk=
=3D0)
0.01597660
Kurtosis           0.55162                         Signif Level (Ku=
=3D0)
0.25977252





---------- End of message ----------

From: Nesrin Okay <okay@boun.edu.tr>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: HELP
Date: Fri, 7 Nov 1997 16:40:22 +0300 (MEST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.30

I AM ATTEMTING TO ESTIMATE A SEMIPARAMETRIC ARCH MODEL.

 

ANY HELP IS APPRECIATED.
 

NESRIN OKAY

ECONOMIST

BOGAZICI UNIVERSITY

ISTANBUL


---------- End of message ----------

From: James Peery Cover <jcover@ua1vm.ua.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Robusterrors option
Date: Fri, 07 Nov 1997 10:46:51 -0600
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.01 [en] (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

If I use the robusterrors option with lags=12 to correct for serial
correlation of unknown form, is the resulting covariance matrix still
heteroscedasticity consistent?
Thank you.

James P. Cover, University of Alabama


---------- End of message ----------

From: "Estima" <estima@estima.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Covariance matrix
Date: Fri, 7 Nov 1997 11:04:10 -0600
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Win32 (v2.54) (via Mercury MTS (Bindery) v1.30)

> 
> I want to compute the kovariance matrix of severel time series.
> Unfortunately rats doesn't compute what I expect.
> Below you see the output of the covariance/correlation matrix. The diagonal
> elements (variances) are completely
> different from the variances I get for the individual time series. Further
> is the covariance and correlation different from what I get with
> Eviews and Excel although I use the same data series.
> 
> Questions:
> 1) What's wrong?
> 2) How can I get a complete kovariance matrix ( without correlations)?
>

I assume you are using the VCV instruction. As you've noted, VCV 
prints out a combination of the covariance and correlation matrices 
(correlations above the diagonal). See Append F, page 1, for details.

If you just want to see the covariance matrix, you can use the MATRIX 
option (which saves only the covariances) and then display that 
matrix:

VCV(MATRIX=V,NOPRINT)
# X Y Z
WRITE V

Also, note that VCV has a CENTERED option which controls whether or
not means are subtracted from the variables (the default is
NOCENTERED). Be sure to use the appropriate option for your
application.

Finally, VCV uses a (1/N) divisor, whereas the STATISTICS
instruction using a (1/N-1), so the results produced by the two
instructions differ by a factor of (N-1)/N (or vice versa, depending
on which way you do the comparison).

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: Axel Schimmelpfennig <Axel.Schimmelpfennig@ifw.uni-kiel.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: ADF-Test for Pooled Data
Date: Sun, 09 Nov 1997 15:56:11 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: The Kiel Institute of World Economics
X-Mailer: Mozilla 4.02 [en] (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear RATS-Users:
Is there a procedure or some other program available that will do a test
for stationarity on pooled data (Levin/Linn)?
Thanks for your help,
Axel Schimmelpfennig

--
Institut fuer Weltwirtschaft (Kiel Institute of World Economics)
Duesternbrooker Weg 120; D-24100 Kiel
phone: ++49/+431/8814-245; fax: -500 or -521
http://www.uni-kiel.de:8080/ifw/

   a satisfied mind [...:]
   a red or yellow apple hanging from a drooping limb,
   behind it deep blue sky, her hand palm up,
   reaching out to the apple but not touching it.
       Ada in Charles Frazier's Cold Mountain



---------- End of message ----------

From: G Mujtaba Mian <mujtabam@gsb.usyd.edu.au>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Unsubscribe
Date: Mon, 10 Nov 1997 08:52:37 +0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 3.0Gold (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Please unsubscribe my name from the list. 
Thank you.



---------- End of message ----------

From: "Gregory, Richard" <RGregory@VEDP.state.va.us>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: RE: HELP
Date: Mon, 10 Nov 1997 09:57:32 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer:  Microsoft Exchange Server Internet Mail Connector Version 4.0.995.52 (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

What kind of help do you need?

Which type of semiparametric ARCH are you attempting to estimate? I
might already have a program for you.


Richard Gregory, Ph.D.
Senior Economist
Virginia Economic Development Partnership
rgregory@vedp.state.va.us


>----------
>From: 	Nesrin Okay[SMTP:okay@boun.edu.tr]
>Sent: 	Friday, November 07, 1997 8:40 AM
>To: 	RATS Discussion List
>Subject: 	HELP
>
>I AM ATTEMTING TO ESTIMATE A SEMIPARAMETRIC ARCH MODEL.
>
> 
>
>ANY HELP IS APPRECIATED.
> 
>
>NESRIN OKAY
>
>ECONOMIST
>
>BOGAZICI UNIVERSITY
>
>ISTANBUL
>
>

---------- End of message ----------

From: Felix.Maag@SBF.unisg.ch
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Quadratic Programming Problem
Date: Tue, 11 Nov 1997 17:36:29 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: text/plain; charset=us-ascii
Content-transfer-encoding: QUOTED-PRINTABLE
X-Mailer: Mercury MTS (Bindery) v1.30

Hi to everyone.

I'm trying to implement Sharps Asset-Class-Factormodel. It is a quadr=
atic
programming problem.

The restrictions are:    1) all coefficients are >=3D0 (is done autom=
atically
by rats)
               2) the coefficients sum to 1 (1 equality constraint)

Working with lqprog(equalities=3D1) gives me a result but the coeffic=
ients
don't sum to 1.

Qestions: 1) Does somebody see the mistake I made?
         2) Has somebody already programmed Sharp's
Asset-Class-Factormodell with rats?


calendar 1988 2 12
allocate 1996:12
open data o:\sbf\diss\ergebnis\brd\rats\daten\rfibrd.rat
data(FOR=3Drats, org=3Dobs)

* declare the relavant vectors and matrices for solving the quadratic
programming problem (three unknowns, 1 equality constraint)
declare rectangular A(1,3) Q(3,3)
declare vector C(3) B(1)

* computes the covariance between the asset return and the indices (v=
ector
c)
vcv(matrix=3DW, centred, noprint)
# rd9 rdi20 rdi50 rdi81
compute C(1) =3D -W(2,1)
compute C(2) =3D -W(3,1)
compute C(3) =3D -W(4,1)

* computes the covariance matrix of the indices (matrix Q)
VCV(MATRIX=3DV,centred,NOPRINT)
# rdi20 rdi50 rdi81
ewise Q(I,J)=3D%if(I=3D=3DJ,V(I,J), V(I,J))

* the three coefficients sum to 1
compute B =3D ||1||

* equality constraint (sum of the three coefficents)
compute A =3D ||1, 1, 1||

lqprog(equalities=3D1) results C A B Q

Result:

QPROG converges at 5 iterations
minimum =3D    -0.00002076
Solution x =3D
 0.00000000000 -0.00000000000  0.28252375204

Thanks for help!!!

Felix Maag, University of St. Gallen, Switzerland





---------- End of message ----------

From: Frederick Bourgoin <fbourgoin@millenniumglobal.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Quadratic Programming Problem
Date: Tue, 11 Nov 1997 18:21:26 +0000
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: MillenniuM Global Investments Ltd.
X-Mailer: Mozilla 4.03 [en] (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit



Felix.Maag@SBF.unisg.ch wrote:

> Hi to everyone.
>
> I'm trying to implement Sharps Asset-Class-Factormodel. It is a quadratic
> programming problem.
>
> The restrictions are:    1) all coefficients are >=0 (is done automatically
> by rats)
>                2) the coefficients sum to 1 (1 equality constraint)
>
> Working with lqprog(equalities=1) gives me a result but the coefficients
> don't sum to 1.
>
> Qestions: 1) Does somebody see the mistake I made?
>          2) Has somebody already programmed Sharp's
> Asset-Class-Factormodell with rats?
>
> calendar 1988 2 12
> allocate 1996:12
> open data o:\sbf\diss\ergebnis\brd\rats\daten\rfibrd.rat
> data(FOR=rats, org=obs)
>
> * declare the relavant vectors and matrices for solving the quadratic
> programming problem (three unknowns, 1 equality constraint)
> declare rectangular A(1,3) Q(3,3)
> declare vector C(3) B(1)
>
> * computes the covariance between the asset return and the indices (vector
> c)
> vcv(matrix=W, centred, noprint)
> # rd9 rdi20 rdi50 rdi81
> compute C(1) = -W(2,1)
> compute C(2) = -W(3,1)
> compute C(3) = -W(4,1)
>
> * computes the covariance matrix of the indices (matrix Q)
> VCV(MATRIX=V,centred,NOPRINT)
> # rdi20 rdi50 rdi81
> ewise Q(I,J)=%if(I==J,V(I,J), V(I,J))
>
> * the three coefficients sum to 1
> compute B = ||1||
>
> * equality constraint (sum of the three coefficents)
> compute A = ||1, 1, 1||
>
> lqprog(equalities=1) results C A B Q
>
> Result:
>
> QPROG converges at 5 iterations
> minimum =    -0.00002076
> Solution x =
>  0.00000000000 -0.00000000000  0.28252375204
>
> Thanks for help!!!
>
> Felix Maag, University of St. Gallen, Switzerland

  I posted the same issue on the user list here 2 weeks ago!!!
It seems that it has do with RATS itself you should use the program that is on
the web site for Quadratic programming and not the one built in, it seems that
there is a bug a Estima is looking at it at the moment (Hopefully)

Frederick Bourgoin
Millennium Global Investments Ltd. London
fbourgoin@compuserve.com



---------- End of message ----------

From: "Estima" <estima@estima.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Quadratic Programming Problem
Date: Tue, 11 Nov 1997 14:30:17 -0600
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Win32 (v2.54) (via Mercury MTS (Bindery) v1.30)


> 
> 
> Felix.Maag@SBF.unisg.ch wrote:
> 
> > Hi to everyone.
> >
> > I'm trying to implement Sharps Asset-Class-Factormodel. It is a quadratic
> > programming problem.
> >
> 
>   I posted the same issue on the user list here 2 weeks ago!!!
> It seems that it has do with RATS itself you should use the program that is on
> the web site for Quadratic programming and not the one built in, it seems that
> there is a bug a Estima is looking at it at the moment (Hopefully)
> 

Dear Frederick et al:

We've located the problem--a last minute change was made to fix
another problem and, unfortunately, a corresponding loop index
wasn't modified correctly. 

The result of this bug is that problems with equality constraints may
fail--one or more of the equality constraints may get "turned off", 
so that the "solution" would not meet the constraints (and thus the 
failure should be fairly obvious).

I have just now finished compiling WinRATS-32 with the fixed source
code, and will be testing it this afternoon. If all goes well, I
should have fixed versions of WinRATS-32 and WinRATS ready by
tomorrow, with RATS386 shortly thereafter. 

So, if you have 4.3 and are using the LQPROG instruction, please 
contact me directly at estima@estima.com.

Sincerely,
Tom Maycock

--
+-----------------------------+-----------------------------------------+
| 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" <RATS-L@efs.mq.edu.au>
Subject: Re: Quadratic Programming Problem
Date: Tue, 11 Nov 1997 15:27:38 -0500 (EST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.30



---------- End of message ----------

From: John Cline <jcline@glinx.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Quadratic Programming Problem
Date: Tue, 11 Nov 1997 16:47:51 -0400
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
X-Mailer: Mercury MTS (Bindery) v1.30


>Dear Frederick et al:
>
>We've located the problem--a last minute change was made to fix
>another problem and, unfortunately, a corresponding loop index
>wasn't modified correctly.
>
>The result of this bug is that problems with equality constraints may
>fail--one or more of the equality constraints may get "turned off",
>so that the "solution" would not meet the constraints (and thus the
>failure should be fairly obvious).
>
>I have just now finished compiling WinRATS-32 with the fixed source
>code, and will be testing it this afternoon. If all goes well, I
>should have fixed versions of WinRATS-32 and WinRATS ready by
>tomorrow, with RATS386 shortly thereafter.
>
>So, if you have 4.3 and are using the LQPROG instruction, please
>contact me directly at estima@estima.com.


Tom,


What about the new Mac versions?

John Cline

Amaranth And Associates &
Amaranth Co-operative Enterprises Ltd.
PO Box 448
Wolfville, Nova Scotia,  B0P 1X0
1-902-542-4002 ( Fax & Voice )
Email:  jcline@glinx.com
Homepage:  http://www.glinx.com/users/jcline/
Worker Owned

Environmental Products, Education, Sustainable Development And Consulting
For A Conserver Society



---------- End of message ----------

From: "Estima" <estima@estima.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Quadratic Programming Problem
Date: Tue, 11 Nov 1997 17:02:44 -0600
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Win32 (v2.54) (via Mercury MTS (Bindery) v1.30)

> 
> Tom,
> 
> 
> What about the new Mac versions?
> 

What about them? I don't mean that flippantly--I'm just not sure what 
you're asking.

Tom Maycock
 
--
+-----------------------------+-----------------------------------------+
| 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: Olaf Hartmann <uog0@rz.uni-karlsruhe.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: GARCH residuals, F-test
Date: Thu, 13 Nov 1997 23:58:23 +-100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Mercury MTS (Bindery) v1.30

Hi,
I estimated a GARCH(1,1)-model and I would like to know,
how I can calculate the residuals of the model:

SET v = 0.0
NONLIN a0 a1 b1
FRML var = a0 + a1*e{1}**2 + b1*w{1}
FRML L = (v = var), -.5*(log(var)+e**2/var)
COMPUTE a0 = ..., a1 = ..., b1 = ...
MAXIMIZE(METHOD=BHHH,ITERATIONS=100) L

After estimating the model I would like to do some F-tests,
using the TEST or the RESTRICT instruction:

TEST
# 3
# 0

or 

RESTRICT 1
# 2 3
# 1 1 1

I am expecting to get the output:

F(1,...)=...

but I always get

Chi-Squared(1)=...

Did I make a mistake in thinking or in programming?

Any suggestions would be appreciated.

Best regards,
Olaf.

----------------------------------------------------
Olaf Hartmann
Institute for Statistics and Econometrics
University of Karlsruhe, Germany


---------- End of message ----------

From: Frieder Knuepling <knueplin@ruf.uni-freiburg.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Logit-Function.
Date: Fri, 14 Nov 1997 12:51:26 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 3.04Gold (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear RATS-users,

I am facing the following problem: 
In an estimation proedure I am using the logit-function
	exp(x)/(1+exp(x))
which attains values between 0 and 1. It happens that the x-value is so
large that the exp-function is not computable for RATS. Then the value
of the logit-funcion is set to NA instead of something slightly less
than one. What should I do?

		Frieder Knuepling

-- 
Frieder Knuepling
Albert-Ludwigs-Universitaet Freiburg
Institut fuer Allgemeine Wirtschaftsforschung
Abteilung Statistik und Oekonometrie
Belfortstr. 24
D-79098 Freiburg
Tel +49 761 / 203 - 2341
Fax +49 761 / 203 - 2298

---------- End of message ----------

From: "Christopher F Baum" <baum@bc.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Logit-Function.
Date: Fri, 14 Nov 1997 07:45:27 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mulberry (MacOS) [1.3.0b5, s/n P020-300786-009] (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


Since all you are saying with very large x is that the cumulative
distribution function is indistinguishable from 1 (or for very small x, from
0) why not just trap the case with, e.g.

set logodds = %if(x<BIGNUM,exp(x)/(1.0+exp(x)),1.0)

where BIGNUM is the largest x which can be exp'd in RATS double precision
implementation? On my UNIX RATS that would seem to be about 700:

?do i=650,750,10
?comp z=exp(float(i))
?dis i z
?enddo
650  1.95620e+282
660  4.30882e+286
670  9.49080e+290
680  2.09049e+295
690  4.60461e+299
700  1.01423e+304
710      NA
720      NA
730      NA
740      NA
750      NA

A similar statement could trap very small values of x. Realistically even
x=50 is going to evaluate to 1 anyway.

Kit Baum


--On Fri, Nov 14, 1997 12:51 PM +0100 "Frieder Knuepling"
<knueplin@ruf.uni-freiburg.de> wrote: 

> Dear RATS-users,
> 
> I am facing the following problem: 
> In an estimation proedure I am using the logit-function
> 	exp(x)/(1+exp(x))
> which attains values between 0 and 1. It happens that the x-value is so
> large that the exp-function is not computable for RATS. Then the value
> of the logit-funcion is set to NA instead of something slightly less
> than one. What should I do?
> 
> 		Frieder Knuepling
> 
> -- 
> Frieder Knuepling
> Albert-Ludwigs-Universitaet Freiburg
> Institut fuer Allgemeine Wirtschaftsforschung
> Abteilung Statistik und Oekonometrie
> Belfortstr. 24
> D-79098 Freiburg
> Tel +49 761 / 203 - 2341
> Fax +49 761 / 203 - 2298



---------------------------------------------------------------------
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: Norman Morin <m1njm00@FRB.GOV>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Logit-Function. 
Date: Fri, 14 Nov 1997 08:07:52 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: exmh version 1.6.9.4 5/23/97 (via Mercury MTS (Bindery) v1.30)
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii

> Dear RATS-users,
> 
> I am facing the following problem: 
> In an estimation proedure I am using the logit-function
> 	exp(x)/(1+exp(x))
> which attains values between 0 and 1. It happens that the x-value is so
> large that the exp-function is not computable for RATS. Then the value
> of the logit-funcion is set to NA instead of something slightly less
> than one. What should I do?
> 
> 		Frieder Knuepling


Frieder,


What might work is the following:

Let e(x) = exp(x).

Since 

e(x)/(1 + e(x))

= ((1 + e(x))/e(x))**-1

= (1/e(x) + 1)**-1

= (e(-x) + 1)**-1,

you could calculate the last expression instead, which should avoid 
the overflow problem.  

I hope this may help.

Norm

-- 

                        Norman J. Morin 
               nmorin@frb.gov or m1njm00@frb.gov  
---------------------------------------------------------------
Industrial Output Section * Division of Research and Statistics
Board of Governors of the Federal Reserve System * Mail Stop 82
            Washington, D.C. 20551 * (202) 452-2476
---------------------------------------------------------------



---------- End of message ----------

From: <kohns@IIW.UNI-BONN.DE>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Logit-Function.
Date: Fri, 14 Nov 1997 14:34:11 MET-1MEST
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: IIW - University of Bonn
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Windows (v2.42a) (via Mercury MTS (Bindery) v1.30)

On 14 Nov 1997, Christopher F Baum wrote:

> 
> Since all you are saying with very large x is that the cumulative
> distribution function is indistinguishable from 1 (or for very small x, from
> 0) why not just trap the case with, e.g.
> 
> set logodds = %if(x<BIGNUM,exp(x)/(1.0+exp(x)),1.0)
> 
> where BIGNUM is the largest x which can be exp'd in RATS double precision
> implementation? On my UNIX RATS that would seem to be about 700:
> 
> ?do i=650,750,10
> ?comp z=exp(float(i))
> ?dis i z
> ?enddo
> 650  1.95620e+282
> 660  4.30882e+286
> 670  9.49080e+290
> 680  2.09049e+295
> 690  4.60461e+299
> 700  1.01423e+304
> 710      NA
> 720      NA
> 730      NA
> 740      NA
> 750      NA
> 
> A similar statement could trap very small values of x. Realistically even
> x=50 is going to evaluate to 1 anyway.
> 
> Kit Baum
> 
An alternative and possibly easier method would be to use the 
equivalent function 1/(exp(-x)+1), which should work fine as long as 
x does not take large (in absolute terms) negative values.

***************************************************
* Dipl.-Volkswirt Stephan Kohns                   *
* Institute for International Economics           *
* University of Bonn                              *
* Germany                                         *
* Tel.: 0049-228-739234                           *
* E-Mail: kohns@iiw.uni-bonn.de                   *
***************************************************

---------- End of message ----------

From: Frieder Knuepling <knueplin@ruf.uni-freiburg.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Logit-Function.
Date: Fri, 14 Nov 1997 15:27:39 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 3.04Gold (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thank you for the suggestions. 
The problem with using 
	1/(exp(-x)+1)
is that this does not work for a large negative x, but Kit Baum's
	%if(x<BIGNUM,exp(x)/(1.0+exp(x)),1.0)
for BIGNUM=700 is exactly what I need.

		Frieder

-- 
Frieder Knuepling
Albert-Ludwigs-Universitaet Freiburg
Institut fuer Allgemeine Wirtschaftsforschung
Abteilung Statistik und Oekonometrie
Belfortstr. 24
D-79098 Freiburg
Tel +49 761 / 203 - 2341
Fax +49 761 / 203 - 2298

---------- End of message ----------

From: Frieder Knuepling <knueplin@ruf.uni-freiburg.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Matrix inversion.
Date: Fri, 14 Nov 1997 15:45:16 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 3.04Gold (Win95; I) (via Mercury MTS (Bindery) v1.30)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear RATS users,

in an estimation pocedure I need to compute the inverse of a
(2x2)-matrix which, from time to time, happens to be non-invertible for
RATS. Instead of the procedure being interupted and the message 
	## MAT14. Non-invertible Matrix...
being displayed I would like to control what happens next in such
situations. Which is the criterion RATS uses to check wether it can
invert a matrix or not? Apparently it is not the determinant, since e.g
the matrix
	 1.2266e-008   -3.0085e-010
	-3.0085e-010    1.4803e-011
with determinant=9.10643e-020 is non-invertible for RATS.

Yours,

		Frieder Knuepling


-- 
Frieder Knuepling
Albert-Ludwigs-Universitaet Freiburg
Institut fuer Allgemeine Wirtschaftsforschung
Abteilung Statistik und Oekonometrie
Belfortstr. 24
D-79098 Freiburg
Tel +49 761 / 203 - 2341
Fax +49 761 / 203 - 2298

---------- End of message ----------