bugs-r-devel

[Bugs-r-devel] R2WinBUGS bug fix

[Dawn W. <dwo...@in...>]

I found a bug in R2WinBUGS, which can be reproduced using the code =
below.  As
a fix, I propose removing a single line of code in bugs.sims:

  sims <- sims [sample(n.sims),]    # scramble (for convenience in =
analysis)

This line of code scrambles the samples for each parameter.  What is the
reason for doing this?  For some reason, when sims is a matrix with a =
single
column (when the user has requested that only one parameter be returned) =
this
line changes sims from a matrix object to a vector object. =20

Thank you,
Dawn



model.file <- function(){
	   for (j in 1:J){
         y[j] ~ dnorm (theta[j], tau.y[j])
         theta[j] ~ dnorm (mu.theta, tau.theta)
         tau.y[j] <- pow(sigma.y[j], -2)
       }
       mu.theta ~ dnorm (0.0, 1.0E-6)
       tau.theta <- pow(sigma.theta, -2)
       sigma.theta ~ dunif (0, 1000)
}


# Some example data (see ?schools for details):
data(schools)

J <- nrow(schools)
y <- schools$estimate
sigma.y <- schools$sd
data <- list ("J", "y", "sigma.y")
inits <- function(){
    list(theta=3Drnorm(J, 0, 100), mu.theta=3Drnorm(1, 0, 100),
         sigma.theta=3Drunif(1, 0, 100))
}
## or alternatively something like:
# inits <- list(
#   list(theta=3Drnorm(J, 0, 90), mu.theta=3Drnorm(1, 0, 90),
#        sigma.theta=3Drunif(1, 0, 90)),
#   list(theta=3Drnorm(J, 0, 100), mu.theta=3Drnorm(1, 0, 100),
#        sigma.theta=3Drunif(1, 0, 100))
#   list(theta=3Drnorm(J, 0, 110), mu.theta=3Drnorm(1, 0, 110),
#        sigma.theta=3Drunif(1, 0, 110)))

parameters <- c("mu.theta")

## Not run:=20
## You may need to edit "bugs.directory",
## also you need write access in the working directory:
schools.sim <- bugs(data, inits, parameters, model.file,
    n.chains=3D1, n.iter=3D5000,
    bugs.directory=3D"c:/Program Files/WinBUGS14/",
    working.directory=3DNULL, clearWD=3DTRUE, DIC =3D F)
print(schools.sim)

Re: [Bugs-r-devel] [Fwd: [R2WinBUGS] bugs]

[Andrew G. <ge...@st...>]

I have only one suggestion.  Jouni's idea is good, but n.sims should be 
the total number of eventual iterations kept.  This is useful for 
post-processing.
Thus, I suggest that Jouni's new parameter should be called 
n.sims.target or n.sims.approx or something like that, to indicate that 
it represents the number of sims that you are trying to get, 
approximately.  Then
n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/n.sims.approx))
and n.sims is defined as already done in R2WinBUGS.
Andrew


>
>
>
> -------- Original Message --------
> Subject: [R2WinBUGS] bugs
> Date: Thu, 20 Sep 2007 12:02:41 +0200
> From: jou...@no...
> To: li...@st...
>
> Hi Uwe
>
> There's a slight bug in R2WinBUGS (2.0.4 and apparently also for 2.1.6).
> The n.thin parameter default value is defined as
>
>         n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/1000)),
>
> which restricts the number of simulations to keep to 1000. We should have
> a parameter
>
>         n.sims = 1000,
>
> and redefine n.thin as
>
>         n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/n.sims)),
>
> to allow changing the number of simulations to keep, easily.
>
>
> [.....]
>
> regards
>
> Jouni

-- 
Andrew Gelman
Professor, Department of Statistics
Professor, Department of Political Science
Director, Applied Statistics Center
Columbia University, New York
ge...@st...
www.stat.columbia.edu/~gelman

Usual schedule:

Mon, Tues, Fri in Statistics Dept ofc:
  Social Work Bldg (Amsterdam Ave at 122 St), Room 1016
  phone 212-851-2142, fax 212-851-2164
Wed, Thurs in Political Science Dept ofc:
  International Affairs Bldg (Amsterdam Ave at 118 St), Room 731
  phone 212-854-7075, fax 212-222-0598
Wed, Thurs from 2:30pm onward in Playroom:
  International Affairs Bldg, Room 707

I'm out of town 11-12 Oct, 26 Oct, and 15-23 Nov.

Re: [Bugs-r-devel] [Fwd: [R2WinBUGS] bugs]

[Ben B. <bo...@zo...>]

  Fine with me.

Uwe Ligges wrote:
> Hi,
>
> here is a proposal by Jouni Kerman. Opinions? Change as proposed?
>
> Uwe
>
>
> -------- Original Message --------
> Subject: [R2WinBUGS] bugs
> Date: Thu, 20 Sep 2007 12:02:41 +0200
> From: jou...@no...
> To: li...@st...
>
> Hi Uwe
>
> There's a slight bug in R2WinBUGS (2.0.4 and apparently also for 2.1.6).
> The n.thin parameter default value is defined as
>
>          n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/1000)),
>
> which restricts the number of simulations to keep to 1000. We should have
> a parameter
>
>          n.sims = 1000,
>
> and redefine n.thin as
>
>          n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/n.sims)),
>
> to allow changing the number of simulations to keep, easily.
>
>
> [.....]
>
> regards
>
> Jouni
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Bugs-r-devel mailing list
> Bug...@li...
> https://lists.sourceforge.net/lists/listinfo/bugs-r-devel
>   

Re: [Bugs-r-devel] [Fwd: [R2WinBUGS] bugs]

[Gregor G. <gre...@bf...>]

Uwe Ligges wrote:
> Hi,
> 
> here is a proposal by Jouni Kerman. Opinions? Change as proposed?

Sounds reasonable for me.

Gregor

[Bugs-r-devel] [Fwd: [R2WinBUGS] bugs]

[Uwe L. <li...@st...>]

Hi,

here is a proposal by Jouni Kerman. Opinions? Change as proposed?

Uwe


-------- Original Message --------
Subject: [R2WinBUGS] bugs
Date: Thu, 20 Sep 2007 12:02:41 +0200
From: jou...@no...
To: li...@st...

Hi Uwe

There's a slight bug in R2WinBUGS (2.0.4 and apparently also for 2.1.6).
The n.thin parameter default value is defined as

         n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/1000)),

which restricts the number of simulations to keep to 1000. We should have
a parameter

         n.sims = 1000,

and redefine n.thin as

         n.thin = max(1, floor(n.chains * (n.iter - n.burnin)/n.sims)),

to allow changing the number of simulations to keep, easily.


[.....]

regards

Jouni

[Bugs-r-devel] SF.net SVN: bugs-r: [74] trunk/R2WinBUGS/R/bugs.run.R

[<li...@us...>]

Revision: 74
          http://bugs-r.svn.sourceforge.net/bugs-r/?rev=74&view=rev
Author:   ligges
Date:     2007-09-25 04:33:02 -0700 (Tue, 25 Sep 2007)

Log Message:
-----------
bugfix: case sensitiveness!

Modified Paths:
--------------
    trunk/R2WinBUGS/R/bugs.run.R

Modified: trunk/R2WinBUGS/R/bugs.run.R
===================================================================
--- trunk/R2WinBUGS/R/bugs.run.R	2007-09-25 11:32:20 UTC (rev 73)
+++ trunk/R2WinBUGS/R/bugs.run.R	2007-09-25 11:33:02 UTC (rev 74)
@@ -49,6 +49,6 @@
   } else {
     tmp <- scan("coda1.txt", character(), sep="\n")
   }
-  if(length(grep("Bugs did not run correctly", tmp)) > 0)
+  if(length(grep("BUGS did not run correctly", tmp)) > 0)
     stop("Look at the log file and\ntry again with 'debug=TRUE' to figure out what went wrong within Bugs.")
 }


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

[Bugs-r-devel] SF.net SVN: bugs-r: [73] trunk/R2WinBUGS/DESCRIPTION

[<li...@us...>]

Revision: 73
          http://bugs-r.svn.sourceforge.net/bugs-r/?rev=73&view=rev
Author:   ligges
Date:     2007-09-25 04:32:20 -0700 (Tue, 25 Sep 2007)

Log Message:
-----------
start new release cycle
take over maintainership from Sibylle

Modified Paths:
--------------
    trunk/R2WinBUGS/DESCRIPTION

Modified: trunk/R2WinBUGS/DESCRIPTION
===================================================================
--- trunk/R2WinBUGS/DESCRIPTION	2007-09-21 08:56:55 UTC (rev 72)
+++ trunk/R2WinBUGS/DESCRIPTION	2007-09-25 11:32:20 UTC (rev 73)
@@ -1,7 +1,7 @@
 Package: R2WinBUGS
 Title: Running WinBUGS and OpenBUGS from R / S-PLUS
-Date: 2007-09-12
-Version: 2.1-6
+Date: 2007-09-25
+Version: 2.1-7
 Author: originally written by Andrew Gelman <ge...@st...>;
  changes and packaged by Sibylle Sturtz <st...@st...>
  and Uwe Ligges <li...@st...>.
@@ -17,6 +17,6 @@
 Suggests: BRugs (>= 0.3-2)
 SystemRequirements: WinBUGS 1.4
 URL: http://www.stat.columbia.edu/~gelman/bugsR/
-Maintainer: Sibylle Sturtz <st...@st...>
+Maintainer: Uwe Ligges <li...@st...>
 License: GPL version 2
 Dialect: R, S-PLUS


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

[Bugs-r-devel] SF.net SVN: bugs-r: [72] trunk/BRugs/R/zzz.R

[<li...@us...>]

Revision: 72
          http://bugs-r.svn.sourceforge.net/bugs-r/?rev=72&view=rev
Author:   ligges
Date:     2007-09-21 01:56:55 -0700 (Fri, 21 Sep 2007)

Log Message:
-----------
OpenBUGS 3.0.3 has been released

Modified Paths:
--------------
    trunk/BRugs/R/zzz.R

Modified: trunk/BRugs/R/zzz.R
===================================================================
--- trunk/BRugs/R/zzz.R	2007-09-20 14:41:52 UTC (rev 71)
+++ trunk/BRugs/R/zzz.R	2007-09-21 08:56:55 UTC (rev 72)
@@ -15,7 +15,7 @@
     }
     
     ".onAttach" <- function(lib, pkg){
-        message("Welcome to BRugs running on OpenBUGS version 3.0.2")
+        message("Welcome to BRugs running on OpenBUGS version 3.0.3")
     }
     
     ".onUnload" <- function(libpath){


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

[Bugs-r-devel] SF.net SVN: bugs-r: [71] trunk/BRugs

[<li...@us...>]

Revision: 71
          http://bugs-r.svn.sourceforge.net/bugs-r/?rev=71&view=rev
Author:   ligges
Date:     2007-09-20 07:41:52 -0700 (Thu, 20 Sep 2007)

Log Message:
-----------
increase version number for next release cycle
add manuals in PDF format

Modified Paths:
--------------
    trunk/BRugs/DESCRIPTION

Added Paths:
-----------
    trunk/BRugs/inst/docs/
    trunk/BRugs/inst/docs/DevMan.pdf
    trunk/BRugs/inst/docs/UserMan.pdf

Modified: trunk/BRugs/DESCRIPTION
===================================================================
--- trunk/BRugs/DESCRIPTION	2007-09-18 14:18:45 UTC (rev 70)
+++ trunk/BRugs/DESCRIPTION	2007-09-20 14:41:52 UTC (rev 71)
@@ -1,7 +1,7 @@
 Package: BRugs
 Title: OpenBUGS and its R / S-PLUS interface BRugs
-Version: 0.4-1
-Date: 2007-09-18
+Version: 0.4-2
+Date: 2007-09-20
 Author: The Chief Software Bug is Andrew Thomas, with web assistance from Real Bug Bob O'Hara. Other members of the BUGS team are statisticians David Spiegelhalter, Nicky Best, Dave Lunn and Ken Rice. Dave Lunn has also made major contributions to the software development. R Code modified, extended and packaged for R by Uwe Ligges and Sibylle Sturtz. Some ideas taken from the R2WinBUGS package based on code by Andrew Gelman.
 Description: An R / S-PLUS package containing OpenBUGS and its R / S-PLUS interface BRugs.
 Maintainer: Uwe Ligges <li...@st...>

Added: trunk/BRugs/inst/docs/DevMan.pdf
===================================================================
(Binary files differ)


Property changes on: trunk/BRugs/inst/docs/DevMan.pdf
___________________________________________________________________
Name: svn:mime-type
   + application/octet-stream

Added: trunk/BRugs/inst/docs/UserMan.pdf
===================================================================
(Binary files differ)


Property changes on: trunk/BRugs/inst/docs/UserMan.pdf
___________________________________________________________________
Name: svn:mime-type
   + application/octet-stream


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

Re: [Bugs-r-devel] MacOS and wine and bugs.directory arguments

[Sparapani, R. <rsp...@mc...>]

One more time ...
env
/usr/bin
which
[1] "no wine in /usr/bin /bin /usr/sbin /sbin /usr/local/bin"
> findUnixBinary(x=3D"winepath")
env
/usr/bin
which
[1] "no winepath in /usr/bin /bin /usr/sbin /sbin /usr/local/bin"

I guess that's where it's barfing.  It's getting "no" as the command =
name.


-----Original Message-----
From: Gregor Gorjanc [mailto:gre...@bf...]
Sent: Wed 9/19/2007 10:38 AM
To: Sparapani, Rodney
Cc: bug...@li...
Subject: Re: MacOS and wine and bugs.directory arguments
=20
Sparapani, Rodney wrote:
>> findUnixBinary(x=3D"wine")
> env
>                                                  WINE
> "/Applications/Darwine/Wine.bundle/Contents/bin/wine"
>> findUnixBinary(x=3D"winepath")
> env
>                                                  WINEPATH
> "/Applications/Darwine/Wine.bundle/Contents/bin/winepath"
>> .Platform$OS.type
> [1] "unix"
>=20

That is weird. When you used

Sys.setenv(WINE=3D...)
Sys.setenv(WINEPATH=3D...)
Sys.setenv(DISPLAY=3D":0.0")
schools.sim <- bugs (data=3Ddata, ...

it worked, when you used

Sys.setenv(DISPLAY=3D":0.0")
schools.sim <- bugs (data=3Ddata, ...

you got an error something like "sh: line 1: no: command not found", but
findUnixBinary(x=3D"wine") and findUnixBinary(x=3D"winepath") works. =
This=20
can not be the case, as findUnixBinary is used when WINE/WINEPATH is=20
NULL. Are you sure your reports are OK? Do you have WINE and WINEPATH=20
variables set by default?

--=20
Lep pozdrav / With regards,
     Gregor Gorjanc
----------------------------------------------------------------------
University of Ljubljana     PhD student
Biotechnical Faculty        www: http://www.bfro.uni-lj.si/MR/ggorjan
Zootechnical Department     blog: http://ggorjan.blogspot.com
Groblje 3                   mail: gregor.gorjanc <at> bfro.uni-lj.si
SI-1230 Domzale             fax: +386 (0)1 72 17 888
Slovenia, Europe            tel: +386 (0)1 72 17 861
----------------------------------------------------------------------

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Dawn W. <dwo...@in...>]

Elsewhere the R2WinBUGS code uses "substring" instead of "substr", which
works in S-PLUS for Windows; we should be able to use it in place of =
substr
in bugs.run, sort.name, winedriveMap and winedriveTr.  The arguments =
"start"
and "stop" in substr are called "first" and "last" in substring.  See =
the
help file in R for more info.

Thanks,
Dawn

-----Original Message-----
From: bug...@li...
[mailto:bug...@li...] On Behalf Of Dawn =
Woodard
Sent: Tuesday, September 18, 2007 10:49 AM
To: gre...@bf...; Uwe Ligges
Cc: bug...@li...
Subject: Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R


That's right, S-PLUS 8.0.4 on Windows does not have a substr function.

Thank you,
Dawn

-----Original Message-----
From: Gregor Gorjanc [mailto:gre...@bf...]=20
Sent: Tuesday, September 18, 2007 11:47 AM
To: Uwe Ligges
Cc: bug...@li...; Dawn Woodard
Subject: Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

Uwe Ligges wrote:
...
>>> Is there a good reason for making this code less secure?
>> S-PLUS, does not have arguments (I do not know why) in substr and it
>> fails with "secure version".
>=20
> I see, thanks! My Windows version of S-PLUS does not have substr() at
> all, hence I thought yours would not have as well. S-PLUS seems to =
have
> a lot differences between operating systems.

Huh :(

Dawn, are our observations correct?

Gregor

-------------------------------------------------------------------------=

This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Bugs-r-devel mailing list
Bug...@li...
https://lists.sourceforge.net/lists/listinfo/bugs-r-devel

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Dawn W. <dwo...@in...>]

I think the corresponding function on S-PLUS for Linux is called "unix".

Let me know whether that works.

Thank you!
Dawn


-----Original Message-----
From: Gorjanc Gregor [mailto:Gre...@bf...]=20
Sent: Wednesday, September 19, 2007 9:53 AM
To: Dawn Woodard; bug...@li...
Subject: RE: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R


> S-PLUS does not have a "system" function on Linux that I am aware of, =
and I
> am looking into why that is.

There is no way for R2WinBUGS to work without system().

Gregor

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Gorjanc G. <Gre...@bf...>]

> S-PLUS does not have a "system" function on Linux that I am aware of, and=
 I
> am looking into why that is.

There is no way for R2WinBUGS to work without system().

Gregor

[Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Dawn W. <dwo...@in...>]

S-PLUS does not have a "system" function on Linux that I am aware of, =
and I
am looking into why that is.

Thanks,
Dawn

-----Original Message-----
From: Gregor Gorjanc [mailto:gre...@bf...]=20
Sent: Wednesday, September 19, 2007 11:39 AM
To: Dawn Woodard
Cc: bug...@li...
Subject: Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

What about system() on Linux?

Dawn Woodard wrote:
> That's right, S-PLUS 8.0.4 on Windows does not have a substr function.
>=20
> Thank you,
> Dawn
>=20
> -----Original Message-----
> From: Gregor Gorjanc [mailto:gre...@bf...]
> Sent: Tuesday, September 18, 2007 11:47 AM
> To: Uwe Ligges
> Cc: bug...@li...; Dawn Woodard
> Subject: Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R
>=20
> Uwe Ligges wrote:
> ...
>>>> Is there a good reason for making this code less secure?
>>> S-PLUS, does not have arguments (I do not know why) in substr and it
>>> fails with "secure version".
>> I see, thanks! My Windows version of S-PLUS does not have substr() at
>> all, hence I thought yours would not have as well. S-PLUS seems to =
have
>> a lot differences between operating systems.
>=20
> Huh :(
>=20
> Dawn, are our observations correct?
>=20
> Gregor


--=20
Lep pozdrav / With regards,
     Gregor Gorjanc
----------------------------------------------------------------------
University of Ljubljana     PhD student
Biotechnical Faculty        www: http://www.bfro.uni-lj.si/MR/ggorjan
Zootechnical Department     blog: http://ggorjan.blogspot.com
Groblje 3                   mail: gregor.gorjanc <at> bfro.uni-lj.si
SI-1230 Domzale             fax: +386 (0)1 72 17 888
Slovenia, Europe            tel: +386 (0)1 72 17 861
----------------------------------------------------------------------

Re: [Bugs-r-devel] MacOS and wine and bugs.directory arguments (was: Recent changes)

[Gorjanc G. <Gre...@bf...>]

> I'm just going to elaborate on /Applications...
>
> GUI (Cocoa/Carbon) apps must be in the /Applications tree or a bad things=
 happen
> (an exception is made for /Developer, but most people don't use/know abou=
t that).
> For any other apps, like command-line tools or X11, Mac OS X doesn't care=
 where you put them
> and it doesn't matter.  Command-line tools are generally in /usr/local un=
less they
> come with a GUI app which means that they are in a bundle like wine/winep=
ath.  So,
> there is no standard place for DarWINE, but most likely it will be found =
in /Applications.
> That's the directory that has the correct permissions for installing syst=
em-wide apps
> and where everything else is going to be put by habit.  Similarly, when y=
ou install
> WinBUGS it tries to place it on your C: drive.  However, if you do that, =
then no one
> else will be able to use it.  So, you choose z:\Applications\WinBUGS14.

OK. Btw, is it /Applications or z:\Applications\?

Now I need to figure out how to know if R is running on a Mac. Will try on =
R-devel.

Gregor

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Gregor G. <gre...@bf...>]

What about system() on Linux?

Dawn Woodard wrote:
> That's right, S-PLUS 8.0.4 on Windows does not have a substr function.
> 
> Thank you,
> Dawn
> 
> -----Original Message-----
> From: Gregor Gorjanc [mailto:gre...@bf...]
> Sent: Tuesday, September 18, 2007 11:47 AM
> To: Uwe Ligges
> Cc: bug...@li...; Dawn Woodard
> Subject: Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R
> 
> Uwe Ligges wrote:
> ...
>>>> Is there a good reason for making this code less secure?
>>> S-PLUS, does not have arguments (I do not know why) in substr and it
>>> fails with "secure version".
>> I see, thanks! My Windows version of S-PLUS does not have substr() at
>> all, hence I thought yours would not have as well. S-PLUS seems to have
>> a lot differences between operating systems.
> 
> Huh :(
> 
> Dawn, are our observations correct?
> 
> Gregor


-- 
Lep pozdrav / With regards,
     Gregor Gorjanc
----------------------------------------------------------------------
University of Ljubljana     PhD student
Biotechnical Faculty        www: http://www.bfro.uni-lj.si/MR/ggorjan
Zootechnical Department     blog: http://ggorjan.blogspot.com
Groblje 3                   mail: gregor.gorjanc <at> bfro.uni-lj.si
SI-1230 Domzale             fax: +386 (0)1 72 17 888
Slovenia, Europe            tel: +386 (0)1 72 17 861
----------------------------------------------------------------------

Re: [Bugs-r-devel] MacOS and wine and bugs.directory arguments

[Gregor G. <gre...@bf...>]

Sparapani, Rodney wrote:
>> findUnixBinary(x="wine")
> env
>                                                  WINE
> "/Applications/Darwine/Wine.bundle/Contents/bin/wine"
>> findUnixBinary(x="winepath")
> env
>                                                  WINEPATH
> "/Applications/Darwine/Wine.bundle/Contents/bin/winepath"
>> .Platform$OS.type
> [1] "unix"
> 

That is weird. When you used

Sys.setenv(WINE=...)
Sys.setenv(WINEPATH=...)
Sys.setenv(DISPLAY=":0.0")
schools.sim <- bugs (data=data, ...

it worked, when you used

Sys.setenv(DISPLAY=":0.0")
schools.sim <- bugs (data=data, ...

you got an error something like "sh: line 1: no: command not found", but
findUnixBinary(x="wine") and findUnixBinary(x="winepath") works. This 
can not be the case, as findUnixBinary is used when WINE/WINEPATH is 
NULL. Are you sure your reports are OK? Do you have WINE and WINEPATH 
variables set by default?

-- 
Lep pozdrav / With regards,
     Gregor Gorjanc
----------------------------------------------------------------------
University of Ljubljana     PhD student
Biotechnical Faculty        www: http://www.bfro.uni-lj.si/MR/ggorjan
Zootechnical Department     blog: http://ggorjan.blogspot.com
Groblje 3                   mail: gregor.gorjanc <at> bfro.uni-lj.si
SI-1230 Domzale             fax: +386 (0)1 72 17 888
Slovenia, Europe            tel: +386 (0)1 72 17 861
----------------------------------------------------------------------

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Dawn W. <dwo...@in...>]

That's right, S-PLUS 8.0.4 on Windows does not have a substr function.

Thank you,
Dawn

-----Original Message-----
From: Gregor Gorjanc [mailto:gre...@bf...]=20
Sent: Tuesday, September 18, 2007 11:47 AM
To: Uwe Ligges
Cc: bug...@li...; Dawn Woodard
Subject: Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

Uwe Ligges wrote:
...
>>> Is there a good reason for making this code less secure?
>> S-PLUS, does not have arguments (I do not know why) in substr and it
>> fails with "secure version".
>=20
> I see, thanks! My Windows version of S-PLUS does not have substr() at
> all, hence I thought yours would not have as well. S-PLUS seems to =
have
> a lot differences between operating systems.

Huh :(

Dawn, are our observations correct?

Gregor

Re: [Bugs-r-devel] MacOS and wine and bugs.directory arguments (was: Recent changes)

[Sparapani, R. <rsp...@mc...>]

> findUnixBinary(x=3D"wine")
env
                                                 WINE=20
"/Applications/Darwine/Wine.bundle/Contents/bin/wine"=20
> findUnixBinary(x=3D"winepath")
env
                                                 WINEPATH=20
"/Applications/Darwine/Wine.bundle/Contents/bin/winepath"=20
> .Platform$OS.type
[1] "unix"
>=20

I'm just going to elaborate on /Applications...

GUI (Cocoa/Carbon) apps must be in the /Applications tree or a bad =
things happen=20
(an exception is made for /Developer, but most people don't use/know =
about that). =20
For any other apps, like command-line tools or X11, Mac OS X doesn't =
care where you put them
and it doesn't matter.  Command-line tools are generally in /usr/local =
unless they
come with a GUI app which means that they are in a bundle like =
wine/winepath.  So,
there is no standard place for DarWINE, but most likely it will be found =
in /Applications.
That's the directory that has the correct permissions for installing =
system-wide apps
and where everything else is going to be put by habit.  Similarly, when =
you install
WinBUGS it tries to place it on your C: drive.  However, if you do that, =
then no one
else will be able to use it.  So, you choose z:\Applications\WinBUGS14.

Rodney

-----Original Message-----
From: Gregor Gorjanc [mailto:gre...@bf...]
Sent: Tue 9/18/2007 6:18 AM
To: Sparapani, Rodney
Cc: bug...@li...
Subject: MacOS and wine and bugs.directory arguments (was: Recent =
changes)
=20
Hi!

Sparapani, Rodney wrote:
>>> Sys.setenv(DISPLAY=3D":0.0")
>>> schools.sim <- bugs (data=3Ddata,
>> +                       inits=3Dinits,
>> +                       parameters.to.save=3Dparameters,
>> +                       model.file=3D"schools.txt",
>> +                       n.chains=3D3,
>> +                       n.iter=3D1000,
>> +   bugs.directory=3D"/Applications/WinBUGS14",
>> +   debug=3DFALSE)
>>
> Doesn't work:
>=20
> sh: line 1: no: command not found
...
> Error in scan(file, what, nmax, sep, dec, quote, skip, nlines, =
na.strings,  :
>         scan() expected 'a real', got 'WinBUGS'

WINE and WINEPATH arguments can be NULL and findUnixBinary() in=20
R/wineutils tries to find corresponding binaries. Can you copy the=20
following (slightly modified findUnixBinary()) into your R. I would like =

to know which command has not been found on mac, is it either which or=20
locate.

findUnixBinary <- function(x)
{
   ## --- Environmental variable ---
   cat("env\n")
   tmp <- Sys.getenv(toupper(x))
   if(nchar(tmp) !=3D 0) return(tmp)
   ## else

   ## --- Standard place ---
   cat("/usr/bin\n")
   tmp <- paste("/usr/bin", x, sep=3D"")
   if(file.exists(tmp)) return(tmp)
   ## else ...

   ## --- Which ---
   cat("which\n")
   tmp <- system(paste("which ", x, sep=3D""), intern=3DTRUE)
   if(length(tmp) !=3D 0) return(tmp)
   ## else ..

   ## --- Locate ---
   cat("locate\n")
   tmp <- system(paste("locate ", x, " | grep bin/", x, "$", sep=3D""),=20
intern=3DTRUE)
   tmp <- tmp[length(tmp)] ## keep only last hit
   if(length(tmp) > 0) return(tmp)

   stop(paste("couldn't find", x, "binary file"))
}

findUnixBinary(x=3D"wine")
findUnixBinary(x=3D"winepath")

> And you are correct.  /Applications is the standard place to install =
GUI applications
> such as R.  Command-line only applications can be installed anywhere.

How do you install WinBUGS on Mac? Do you choose /Aplications folder or =
...

I just want to be sure that /Applications is a safe default as=20
C:\Program Files is on MS Windows.

Additionally, how can we know if R is running on Mac? I guess that

.Platform$OS.type has value "unix" as for Linux/Unix.

--=20
Lep pozdrav / With regards,
     Gregor Gorjanc
----------------------------------------------------------------------
University of Ljubljana     PhD student
Biotechnical Faculty        www: http://www.bfro.uni-lj.si/MR/ggorjan
Zootechnical Department     blog: http://ggorjan.blogspot.com
Groblje 3                   mail: gregor.gorjanc <at> bfro.uni-lj.si
SI-1230 Domzale             fax: +386 (0)1 72 17 888
Slovenia, Europe            tel: +386 (0)1 72 17 861
----------------------------------------------------------------------

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [70] trunk/BRugs/DESCRIPTION

[Uwe L. <li...@st...>]

Folks,

Andrew found out he cannot yet manage to compile a brugs.so that works 
reliable on (my version of) Linux for now, but maybe in the future.
Thanks to Andrew for all his work on OpenBUGS 3.0.2 and thanks for his 
tries to get it working under Linux!!!!

I will release BRugs 0.4-1 with OpenBUGS 3.0.2 (including a slightly 
updated BRugs.dll without need for libtaucs.dll) now and submit to CRAN 
within a couple of minutes.

Best wishes,
Uwe




li...@us... wrote:
> Revision: 70
>           http://bugs-r.svn.sourceforge.net/bugs-r/?rev=70&view=rev
> Author:   ligges
> Date:     2007-09-18 07:18:45 -0700 (Tue, 18 Sep 2007)
> 
> Log Message:
> -----------
> preparing for CRAN release
> 
> Modified Paths:
> --------------
>     trunk/BRugs/DESCRIPTION
> 
> Modified: trunk/BRugs/DESCRIPTION
> ===================================================================
> --- trunk/BRugs/DESCRIPTION	2007-09-18 14:09:25 UTC (rev 69)
> +++ trunk/BRugs/DESCRIPTION	2007-09-18 14:18:45 UTC (rev 70)
> @@ -1,7 +1,7 @@
>  Package: BRugs
>  Title: OpenBUGS and its R / S-PLUS interface BRugs
>  Version: 0.4-1
> -Date: 2007-09-14
> +Date: 2007-09-18
>  Author: The Chief Software Bug is Andrew Thomas, with web assistance from Real Bug Bob O'Hara. Other members of the BUGS team are statisticians David Spiegelhalter, Nicky Best, Dave Lunn and Ken Rice. Dave Lunn has also made major contributions to the software development. R Code modified, extended and packaged for R by Uwe Ligges and Sibylle Sturtz. Some ideas taken from the R2WinBUGS package based on code by Andrew Gelman.
>  Description: An R / S-PLUS package containing OpenBUGS and its R / S-PLUS interface BRugs.
>  Maintainer: Uwe Ligges <li...@st...>
> 
> 
> This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Bugs-r-devel mailing list
> Bug...@li...
> https://lists.sourceforge.net/lists/listinfo/bugs-r-devel

[Bugs-r-devel] SF.net SVN: bugs-r: [70] trunk/BRugs/DESCRIPTION

[<li...@us...>]

Revision: 70
          http://bugs-r.svn.sourceforge.net/bugs-r/?rev=70&view=rev
Author:   ligges
Date:     2007-09-18 07:18:45 -0700 (Tue, 18 Sep 2007)

Log Message:
-----------
preparing for CRAN release

Modified Paths:
--------------
    trunk/BRugs/DESCRIPTION

Modified: trunk/BRugs/DESCRIPTION
===================================================================
--- trunk/BRugs/DESCRIPTION	2007-09-18 14:09:25 UTC (rev 69)
+++ trunk/BRugs/DESCRIPTION	2007-09-18 14:18:45 UTC (rev 70)
@@ -1,7 +1,7 @@
 Package: BRugs
 Title: OpenBUGS and its R / S-PLUS interface BRugs
 Version: 0.4-1
-Date: 2007-09-14
+Date: 2007-09-18
 Author: The Chief Software Bug is Andrew Thomas, with web assistance from Real Bug Bob O'Hara. Other members of the BUGS team are statisticians David Spiegelhalter, Nicky Best, Dave Lunn and Ken Rice. Dave Lunn has also made major contributions to the software development. R Code modified, extended and packaged for R by Uwe Ligges and Sibylle Sturtz. Some ideas taken from the R2WinBUGS package based on code by Andrew Gelman.
 Description: An R / S-PLUS package containing OpenBUGS and its R / S-PLUS interface BRugs.
 Maintainer: Uwe Ligges <li...@st...>


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

[Bugs-r-devel] SF.net SVN: bugs-r: [69] trunk/R2WinBUGS/R/bugs.R

[<li...@us...>]

Revision: 69
          http://bugs-r.svn.sourceforge.net/bugs-r/?rev=69&view=rev
Author:   ligges
Date:     2007-09-18 07:09:25 -0700 (Tue, 18 Sep 2007)

Log Message:
-----------
model.file also under Linux: simplified, more efficient, more secure

Modified Paths:
--------------
    trunk/R2WinBUGS/R/bugs.R

Modified: trunk/R2WinBUGS/R/bugs.R
===================================================================
--- trunk/R2WinBUGS/R/bugs.R	2007-09-18 13:23:04 UTC (rev 68)
+++ trunk/R2WinBUGS/R/bugs.R	2007-09-18 14:09:25 UTC (rev 69)
@@ -37,12 +37,13 @@
     on.exit(setwd(savedWD))
   }
   if(is.function(model.file)){
+      temp <- tempfile("model")
       temp <-
-        ifelse(is.R(),
-               paste(tempfile("model"), "txt", sep="."),
-               ifelse(.Platform$OS.type != "unix",
-                      gsub(".tmp$", ".txt", tempfile("model")),
-                      paste(tempfile("model"), "txt", sep=".")))
+        if(is.R() || .Platform$OS.type != "windows"){
+               paste(temp, "txt", sep=".")
+        } else {
+               gsub("\\.tmp$", ".txt", temp)
+        }
       write.model(model.file, con=temp)
       model.file <- gsub("\\\\", "/", temp)
       if(!is.R()) on.exit(file.remove(model.file), add=TRUE)


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Gregor G. <gre...@bf...>]

Uwe Ligges wrote:
...
>>> Is there a good reason for making this code less secure?
>> S-PLUS, does not have arguments (I do not know why) in substr and it
>> fails with "secure version".
> 
> I see, thanks! My Windows version of S-PLUS does not have substr() at
> all, hence I thought yours would not have as well. S-PLUS seems to have
> a lot differences between operating systems.

Huh :(

Dawn, are our observations correct?

Gregor

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Uwe L. <li...@st...>]

Gregor Gorjanc wrote:
> Uwe Ligges wrote:
> ...
>>
>> Thank you for the update!
>>
>> Unfortunately, this are inappropriate usages of ifelse(): We know that
>> we are working with a length 1 logical vectors here. Using ifelse() is
>> more insecure and inefficient in such a case!
>>
>> I'll change that myself to if(){} else{}.
> 
> I used ifelse to avoid clutter, but if(){} else{} is also Ok for me. 
> Thanks!
> 
>>> -  if(useWINE && (substr(bugs.directory, start=2, stop=2) == ":")) {
>>> +  if(useWINE && (substr(bugs.directory, 2, 2) == ":")) {
>>
>> Is there a good reason for making this code less secure?
> 
> S-PLUS, does not have arguments (I do not know why) in substr and it 
> fails with "secure version".

I see, thanks! My Windows version of S-PLUS does not have substr() at 
all, hence I thought yours would not have as well. S-PLUS seems to have 
a lot differences between operating systems.

Uwe




> 
> Gregor

Re: [Bugs-r-devel] SF.net SVN: bugs-r: [66] trunk/R2WinBUGS/R

[Gregor G. <gre...@bf...>]

Uwe Ligges wrote:
...
> 
> Thank you for the update!
> 
> Unfortunately, this are inappropriate usages of ifelse(): We know that
> we are working with a length 1 logical vectors here. Using ifelse() is
> more insecure and inefficient in such a case!
> 
> I'll change that myself to if(){} else{}.

I used ifelse to avoid clutter, but if(){} else{} is also Ok for me. Thanks!

>> -  if(useWINE && (substr(bugs.directory, start=2, stop=2) == ":")) {
>> +  if(useWINE && (substr(bugs.directory, 2, 2) == ":")) {
> 
> Is there a good reason for making this code less secure?

S-PLUS, does not have arguments (I do not know why) in substr and it 
fails with "secure version".

Gregor