Search String: Display: Description: Sort:

Results:

References: [ +from:mconklin@customresearch.com: 35 ]

Total 35 documents matching your query.

1. Rpart plotting problem (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Thu, 10 Apr 2003 11:46:40 -0500
t
/archives/html/s-news/2003-04/msg00069.html (7,707 bytes)

2. Re: .Fortran call of DLL causing Splus exit (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Fri, 16 Nov 2001 12:19:59 -0600
I sent a reply yesterday but I am not sure if it went through. However, today I am running into a similar problem. When you get a stack overflow in the dll Splus simply exits. This happens because yo
/archives/html/s-news/2001-11/msg00150.html (12,890 bytes)

3. Menu Problems under Windows 2000 (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Fri, 05 Oct 2001 12:54:51 -0500
We recently upgraded(?) our machines from WinNT SP4 to Windows 2000. We have invested a lot of time (and therefore money) in creating GUI menus for Splus 2000. We are finding that the our menus are b
/archives/html/s-news/2001-10/msg00044.html (7,556 bytes)

4. Re: Menu Problems under Windows 2000 (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Fri, 05 Oct 2001 16:26:13 -0500
Sam Buttrey suggested the "deleting the prefs directory", but this did not solve the problem. Any other suggestions or ideas are most welcome. Michael Conklin This may not do the trick, but I found a
/archives/html/s-news/2001-10/msg00048.html (9,288 bytes)

5. rpart bug (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Tue, 04 Sep 2001 10:01:12 -0500
I was running some analyses this weekend using the Rpart code at the MASS 3 site that was so kindly ported for Splus 6 for Windows by Brian Ripley. There appears to be a change in the residuals.rpart
/archives/html/s-news/2001-09/msg00012.html (7,486 bytes)

6. Factor Scores and oblimin rotation (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Thu, 20 Sep 2001 12:02:09 -0500
Dear List: When I use the factanal procedure with an oblique rotation (specifically oblimin), the output shows a correlation matrix between the factors. However, if I extract the factor scores from t
/archives/html/s-news/2001-09/msg00192.html (7,173 bytes)

7. Re: Installing S-Plus version 3.3 in Windows Millenium. (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Fri, 28 Sep 2001 08:53:55 -0500
p
/archives/html/s-news/2001-09/msg00249.html (10,328 bytes)

8. Re: Stack overflow (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Thu, 21 Jun 2001 07:45:06 -0500
You don't say what system you are using, but my experience on Windows (95 and NT) is that when FORTRAN is called from Splus there is not enough stack space for FORTRAN to allocate large data structur
/archives/html/s-news/2001-06/msg00192.html (8,762 bytes)

9. Re: dll.load problem (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Fri, 25 May 2001 07:56:51 -0500
I believe your symbol should be bmlogit=_bmlogit@4*numparams for example if you pass 8 parameters to bmlogit then the symbol would be symbols=list(bmlogit=_BMLOGIT@32) By the way - one thing I really
/archives/html/s-news/2001-05/msg00278.html (8,490 bytes)

10. Re: dll.load with Fortran (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Fri, 25 May 2001 16:41:00 -0500
Try the following: dll.load("C:\MSDEV\Projects\testlink\Debug\testlink.dll", symbols=list(arsim=_ARSIM@12)) The symbols are exported with a leading underscore and a 4*number of parameters indicator.
/archives/html/s-news/2001-05/msg00296.html (11,577 bytes)

11. Re: Bagging & Boosting (Re: Regression Trees) (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Thu, 05 Apr 2001 12:30:11 -0500
A good place to look for information on boosting is www.boosting.org I suspect that you are finding few references because it is primarily in the machine learning literature. Michael Conklin ** Micha
/archives/html/s-news/2001-04/msg00061.html (9,247 bytes)

12. Problems Using Fuzzy Clustering (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Wed, 20 Dec 2000 15:02:29 -0600
I am trying to use the fanny function or MenuFanny in Splus 2000 on Windows NT. It seems that no matter how many clusters I ask for I get 2 out with all clusters having equal probability. Has anyone
/archives/html/s-news/2000-12/msg00162.html (6,615 bytes)

13. [S] Any bugs in AIMD library? (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Tue, 05 Sep 2000 14:34:27 -0500
Has anyone encountered implausible results using Joe Schafer's AIMD library? I was trying to use the program to generate a what I thought was a simple task. I have a series of 1 to 5 scales (40 in al
/archives/html/s-news/2000-09/msg00027.html (7,570 bytes)

14. [S] Change in na.include? (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Wed, 26 Jul 2000 09:10:10 -0500
Has there been a change in the behavior of na.include with Splus 2000. I have a function where I want NAs to be present in the data.frame i.e. not as additional levels of a factor. Previously this fu
/archives/html/s-news/2000-07/msg00190.html (7,391 bytes)

15. [S] weights in glm's (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Wed, 21 Jun 2000 15:56:01 -0500
Dear S experts: I am confused by the help file on glm and gam. Is it appropriate to use case weights in fitting these models or do the weights refer to something else. I would like to use the functio
/archives/html/s-news/2000-06/msg00150.html (7,418 bytes)

16. [S] SUMMARY: weights in glm (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Thu, 22 Jun 2000 08:09:23 -0500
Thanks to Tim Hesterberg and Brian Ripley for their prompt responses. As Tim noted, a good way to see how the program works is to try it out with some known data. What I wanted to know was if the wei
/archives/html/s-news/2000-06/msg00160.html (8,091 bytes)

17. [S] Re: Collinearity Diagnostics (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Wed, 19 Apr 2000 08:38:43 -0500
VIF can be calculated easily as follows: assume lm.obj is your linear regression object VIF<-diag(solve(cor(model.matrix(lm.obj)[,-1]))) model.matrix extracts the predictor matrix. You want to leave
/archives/html/s-news/2000-04/msg00172.html (7,445 bytes)

18. [S] ODBC Corrupting Data (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Mon, 6 Mar 2000 10:08:52 -0600
I am trying to read in a table from a MS Access 97 database and I am having some SERIOUS problems. Specifically - I have a field that is mostly empty CUSTNAME2 which is used for a second name (e.g. a
/archives/html/s-news/2000-03/msg00058.html (7,709 bytes)

19. Re: [S] make matrix symmetric (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Mon, 13 Mar 2000 16:43:57 -0600
Not very elegant but try this. [,1] [,2] [,3] [,4] [1,] 1 5 9 13 [2,] 2 6 10 14 [3,] 3 7 11 15 [4,] 4 8 12 16 x[matrix( c(row(x)[lower.tri(x)], col(x)[lower.tri(x)]), ncol = 2)] [,1] [,2] [,3] [,4] [
/archives/html/s-news/2000-03/msg00127.html (8,931 bytes)

20. [S] Summary: ODBC Corrupting Data (score: 1)
Author: "Michael Conklin" <mconklin@customresearch.com>
Date: Thu, 23 Mar 2000 15:38:15 -0600
I posted this a couple of weeks ago and here is a summary of the Paul Chartraw suggested the following: Michael - Change "Select ..., CUSTNAME2, .... from ....." to "Select ..., IIF(ISNull(CUSTNAME2)
/archives/html/s-news/2000-03/msg00220.html (9,533 bytes)

Current List: 1 - 20
Page: [1] [2]
This search system is powered by Namazu