Search String: Display: Description: Sort:

Results:

References: [ +from:ypeng@math.mun.ca: 40 ]

Total 40 documents matching your query.

1. Combining shared libraries (*.so files) directly? (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Tue, 18 Feb 2003 16:50:53 -0330
This is an off topic. Is it possible to combine a few shared library files (.so files) without changing source files of the libraries? I am not sure whether it makes sense or not. I have a few shared
/archives/html/s-news/2003-02/msg00113.html (7,013 bytes)

2. Draw ellipses in S-PLUS or R? (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Tue, 22 Oct 2002 16:59:52 -0230
Dear S-PLUS/R users: Do you know any default function or a user contributed function that can draw an ellipse with given axes and origin? Thanks for any help. Paul. -- Romance, like alcohol, should b
/archives/html/s-news/2002-10/msg00171.html (6,845 bytes)

3. Re: Draw ellipses in S-PLUS or R? (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Tue, 22 Oct 2002 17:33:28 -0230
Many thanks to Andy Liaw for quickly reminding me of "ellipse" package in CRAN. It is apparently what I wanted. Not sure whether it will work in S-PLUS or not. But I will give it a try. Thank you. Pa
/archives/html/s-news/2002-10/msg00172.html (9,068 bytes)

4. A weird problem of Winhelp in Sqpe (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Thu, 19 Sep 2002 15:00:14 -0230
Dear S-PLUS users: I recently experienced a weird problem of Winhelp in sqpe: If I start S-PLUS 2000 in a Win98se box with sqpe in a dos window or in a Emacs+ESS window and use ?function.name or help
/archives/html/s-news/2002-09/msg00136.html (6,749 bytes)

5. Tenure track positions in Statistics (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Wed, 21 Aug 2002 17:41:53 -0230
This email is forwarded on behalf of Dr. Herbert Gaskill. Please do not reply to me but direct enquiries to Ms Wanda Heath <wanda@math.mun.ca> or Dr. Herbert Gaskill <head@math.mun.ca>. Apologies for
/archives/html/s-news/2002-08/msg00128.html (9,370 bytes)

6. Re: split-population model (cure model) (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Tue, 11 Jun 2002 16:54:14 -0230
Sorry for this late response. I have been away for a while, and only received this message today. I have an S-PLUS package GFCURE, which can do your job. It fits parametric cure models with many dist
/archives/html/s-news/2002-06/msg00088.html (8,685 bytes)

7. Multiple-page postscript device. (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Mon, 05 Nov 2001 13:13:23 -0330
I am curious whether it is possible to start a postscript graphics device in S-PLUS that can produce a multiple-page postscript file. If the built-in postscript() cannot do this, is there any add-on
/archives/html/s-news/2001-11/msg00036.html (6,455 bytes)

8. SUMMARY: Multiple-page postscript device. (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Tue, 06 Nov 2001 12:19:19 -0330
Dear Greg, Thank you so much for your helpful reply. I played with postscript() again and found that you're right. I thought the ps file was not created properly when I saw warning messages from Gsvi
/archives/html/s-news/2001-11/msg00049.html (9,144 bytes)

9. Re: sqpe documentation (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Fri, 09 Nov 2001 13:11:13 -0330
I have used sqpe all the time. But I can't remember exactly where it is documented in the manuals. I checked my autoexec.bat and believed that the following two lines make sqpe.exe work: SET ALWAYS_P
/archives/html/s-news/2001-11/msg00080.html (9,921 bytes)

10. DLL_LOAD in making DLLs for S-PLUS 2000 (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Sun, 07 Oct 2001 11:24:29 -0230
I used to create the DLLs in VC++6.0 by following the instructions in the Programmer's manual and defined DLL_LOAD during compilation. Recently, I tried to use a different name, say FILE_DLL instead
/archives/html/s-news/2001-10/msg00052.html (7,331 bytes)

11. Re: DLL_LOAD in making DLLs for S-PLUS 2000 (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Sun, 07 Oct 2001 15:19:11 -0230
Many thanks to Brian Ripley and Hong Ooi for their quick replies. Their comments, particular those from Brian, make me believe that it must be my fault, and indeed it is. I forgot to change DLL_LOAD
/archives/html/s-news/2001-10/msg00055.html (7,243 bytes)

12. familty = poisson(ink = identity) in glm (score: 1)
Author:
Date: Fri, 17 Aug 2001 13:24:35 -0230
Dear S-PLUS users, The S-PLUS manual shows that family = poisson(link = identity) is legitimate in glm. However, following is what I got Error in .Fortran("glmfit",: subroutine glmfit: 62 Inf value(s
/archives/html/s-news/2001-08/msg00133.html (7,828 bytes)

13. Re: familty = poisson(ink = identity) in glm (score: 1)
Author: y@stats.ox.ac.uk>
Date: Fri, 17 Aug 2001 18:18:25 -0230
Many thanks to Brian for his suggestion. I tried the two methods above, and they all made glm run without errors, but with warning messages. For example, my modified glm.links[, "identity"]$initializ
/archives/html/s-news/2001-08/msg00137.html (9,702 bytes)

14. Re: familty = poisson(ink = identity) in glm (score: 1)
Author: yang@NRCan.gc.ca>
Date: Wed, 22 Aug 2001 00:12:17 -0230
Many thanks to Terry Therneau and Brian Ripley for their further suggestions which solved my questions. Both methods yield results similar to those reported in Agresti's book except that the s.e. of
/archives/html/s-news/2001-08/msg00158.html (7,471 bytes)

15. Maximum Memory Size in S-PLUS 2000 (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Fri, 04 May 2001 15:58:30 -0230
Dear S-PLUS users or developers, I know that S-PLUS 2000 comes with a default value of 2147483647 for options()$memory. My question is: Can I increase this default value to, say, 10GB (Suppose that I
/archives/html/s-news/2001-05/msg00046.html (7,503 bytes)

16. Re: Maximum Memory Size in S-PLUS 2000 (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Mon, 07 May 2001 11:35:51 -0230
I would like to thank Brian Ripley and James Holtman for their quick and definitive answer to my question. Brian reminds me that 2GB should be enough to 75MB data set. But I simply called coxph in S-
/archives/html/s-news/2001-05/msg00062.html (11,221 bytes)

17. Baseline Survival in PH Model for Data in Counting Process Style. (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Wed, 09 May 2001 15:12:25 -0230
Dear S-PLUS users, This is an off-topic question, but I think some of you may know the answer. In SAS, is there any way to get baseline survival estimates from PHREG when survival data are presented
/archives/html/s-news/2001-05/msg00077.html (7,171 bytes)

18. Re: question about mixture models in survival analysis (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Mon, 14 May 2001 11:19:14 -0230
I wrote an S-PLUS/C program package "gfcure" to fit survival data with parametric cure models. It is available from StatLib (http://lib.stat.cmu.edu/DOS/S/). HTH. Paul. -- ___________________________
/archives/html/s-news/2001-05/msg00113.html (8,443 bytes)

19. Penalized Cox models in S-PLUS 5. (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Fri, 02 Mar 2001 17:09:05 -0330
Dear S-PLUS users, When I tried the example given in S-PLUS 2000 Statistics Guide Vol 2, pp. 293-295 in S-PLUS 2000, it worked well as follows: ovarian) Call: coxph(formula = Surv(futime, fustat) ~ r
/archives/html/s-news/2001-03/msg00023.html (7,722 bytes)

20. Survival estimates from coxph() with strata() (score: 1)
Author: "Paul Y. Peng" <ypeng@math.mun.ca>
Date: Sat, 13 Jan 2001 00:47:46 -0330
I found survfit.coxph always returns survival estimates for all strata if strata() appears in the call to coxph, even when newdata is given in survfit.coxph. I suppose that any row in newdata only co
/archives/html/s-news/2001-01/msg00093.html (7,487 bytes)


This search system is powered by Namazu