Mail Archive Search: <+from:vaida@sdac.harvard.edu>

Results:

References: [ +from:vaida@sdac.harvard.edu: 15 ]

Total 15 documents matching your query.

1. Acknowledgments: testing for ANOVA (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Mon, 11 Jun 2001 22:09:46 -0400 (EDT); Thanks to Steve McKinney, Dennis Murphy, Andrew Sinclair, Andy Liaw, Mat Souckup and Greg Arnold for sharing their insight and expertise regarding my ANOVA question. I found particularly useful Denni; /archives/html/s-news/2001-06/msg00093.html (9,297 bytes)
2. testing for ANOVA (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Thu, 24 May 2001 14:17:36 -0400 (EDT); Dear S-newsers, I have a straightforward ANOVA question. Consider a 2-way ANOVA model, such as fit.2way <- lm(y ~ treatment + nnrti, data=actg) Running the anova() command I get: Analysis of Variance; /archives/html/s-news/2001-05/msg00264.html (7,785 bytes)
3. Re: Generating random real numbers (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Wed, 4 Apr 2001 09:30:41 -0400 (EDT); Dear all, 1. Let me open the question a bit: how to sample from a truncated MULTIVARIATE normal, with known parameters. I.e., given MVN(m,V) where m is a p-vector and V is the pxp variance matrix, an; /archives/html/s-news/2001-04/msg00038.html (11,462 bytes)
4. multivariate analysis (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Fri, 23 Mar 2001 15:22:15 -0500 (EST); Dear all, I have a statistical question. Suppose that I collect data on N individuals, on k items. E.g., how well individual i answers question (i) assessing the strength of correlation between the k; /archives/html/s-news/2001-03/msg00190.html (7,551 bytes)
5. Re: Biting the hand that feeds you (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Wed, 7 Feb 2001 10:48:28 -0500 (EST); As a member of the S-news list, I suggest that we save this for what is intended, which in my understanding is a forum for learning and sharing expertise in S-related and statistical issues. I'm conv; /archives/html/s-news/2001-02/msg00067.html (9,028 bytes)
6. RE: Data is or data are? (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Fri, 10 Nov 2000 14:31:35 -0500 (EST); This from http://www.encyclopedia.com/articles/13717.html Webster, Daniel 1782-1852, American politician; b. Salisbury (now in Franklin), N.H. A lawyer famous for his oratory, Webster won national re; /archives/html/s-news/2000-11/msg00098.html (12,176 bytes)
7. Re: stepwise coxph? (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Mon, 23 Oct 2000 13:51:02 -0400 (EDT); I agree with prof. Ripley that the partial log-likelihood is the thing to use, but I'd like to point out that this is the same as the (full) log-likelihood! Indeed, the partial likelihood is in fact; /archives/html/s-news/2000-10/msg00201.html (9,234 bytes)
8. [S] coxph() (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Tue, 19 Sep 2000 17:51:55 -0400 (EDT); Dear S-newsers, I have a problem for coxph() buffs. The baseline hazard is returned by coxph.detail(fit), where fit is a coxph() object. When fitting the model with an offset in the linear term, the; /archives/html/s-news/2000-09/msg00135.html (8,100 bytes)
9. [S] fisher.test (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Thu, 10 Aug 2000 23:22:03 -0400 (EDT); edict.rpart use fit$splits as parameter instead of fit$frame$splits.( Splus function predict.tree use fit$frame$splits). If I split one leaf of fit to gene; /archives/html/s-news/2000-08/msg00104.html (7,759 bytes)
10. RE: [S] fisher.test (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Fri, 11 Aug 2000 12:32:35 -0400 (EDT); records all bugs associated with fisher.test have been fixed (other than an open enhancement request for 1-sided as well as 2-sided p-values). This applie; /archives/html/s-news/2000-08/msg00116.html (11,668 bytes)
11. RE: [S] fisher.test (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Fri, 11 Aug 2000 12:49:45 -0400 (EDT); 1 for Unix. This is the sequence of commands that raised my suspicions: fisher.test( matrix(c(1,47,9,41), ncol=2) ) fisher.test( matrix(c(1,48,9,41), ncol; /archives/html/s-news/2000-08/msg00117.html (13,839 bytes)
12. RE: [S] fisher.test (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Fri, 11 Aug 2000 15:02:15 -0400 (EDT); ions since I made my post earlier today about PowerPoint and wmf files. Here are a few more ramblings, to answer some of the questions I've been getting. 1; /archives/html/s-news/2000-08/msg00120.html (8,259 bytes)
13. [S] dataframe subset (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Tue, 15 Aug 2000 13:37:04 -0400 (EDT); . I would like the default for the tick placement to be "in" rather than "out". I suspect that I can use the guiModify function for this task but do not un; /archives/html/s-news/2000-08/msg00147.html (7,624 bytes)
14. [S] proportional odds model (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Wed, 26 Jul 2000 23:29:27 -0400 (EDT); Dear S-plusers, 1. Is anyone aware of s-plus implementations for proportional odds models for categorical data? (I believe there is something available in SAS; I've used before PLUM written by Peter; /archives/html/s-news/2000-07/msg00208.html (7,605 bytes)
15. Re: [S] proportional odds model (score: 1): Author: Florin Vaida <vaida@sdac.harvard.edu>; Date: Mon, 31 Jul 2000 22:13:02 -0400 (EDT); Thanks to AJ Rossini, B Ripley and F Harrell Jr for indicating the functions polr (Mass) and lrm (Design) for the fixed effects model, and to F Harrell again and Tim Johnson for reference to stand-al; /archives/html/s-news/2000-07/msg00244.html (7,869 bytes)

Current List: 1 - 15
Page: [1]