Mail Archive Search: <+from:zaslavsk@hcp.med.harvard.edu>

Results:

References: [ +from:zaslavsk@hcp.med.harvard.edu: 107 ]

Total 107 documents matching your query.

1. Re: matrix "multiplication" (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Fri, 8 Jul 2005 18:13:14 -0400 (EDT); Just use ordinary multiplication, C = A*B --0-2010184243-1120858920=:40489 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Dear Friends: I need to perform the matrix mult; /archives/html/s-news/2005-07/msg00045.html (8,572 bytes)
2. Re: Complex sample designs in S-plus (and R) (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Fri, 3 Jun 2005 22:59:14 -0400 (EDT); Yes, see links to Tom Lumley's R package (believed to work with minor modifications at most in S-Plus) at http://www.hcp.med.harvard.edu/statistics/survey-soft/r.html; /archives/html/s-news/2005-06/msg00010.html (7,843 bytes)
3. Re: last question of the week (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Wed, 11 May 2005 19:37:09 -0400 (EDT); -; /archives/html/s-news/2005-05/msg00057.html (9,474 bytes)
4. Re: Auto point numbering with plot(x,y) (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Sun, 15 May 2005 14:11:19 -0400 (EDT); k; /archives/html/s-news/2005-05/msg00091.html (8,019 bytes)
5. Re: Padding out a vector of characters to a constant width (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Mon, 16 May 2005 13:30:24 -0400 (EDT); t; /archives/html/s-news/2005-05/msg00094.html (8,113 bytes)
6. Re: conditional or paired logisitic regression (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Thu, 3 Mar 2005 15:56:15 -0500 (EST); Conditional logistic regression is equivalent to a stratified Cox proportional hazards model, if you set up the data in the right way. Hence in various software, e.g. SAS, S-Plus, you can use PH soft; /archives/html/s-news/2005-03/msg00021.html (7,359 bytes)
7. Re: Sudaan? (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Wed, 16 Mar 2005 11:55:38 -0500 (EST); The nearest thing is the "survey" package by Thomas Lumley of U of Washington Biostatistics. The package was written for R but Thomas has kindly included some hints for porting it to S (a nontrivial; /archives/html/s-news/2005-03/msg00066.html (7,408 bytes)
8. Re: complex survey analyses (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Sun, 27 Feb 2005 18:46:54 -0500 (EST); There is a package of routines written for R for this purpose. See manual http://finzi.psych.upenn.edu/R/library/survey/html/00Index.html I couldn't tell you for sure whether this works in S-Plus wit; /archives/html/s-news/2005-02/msg00156.html (8,109 bytes)
9. Re: order in browser mode (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Sat, 29 Jan 2005 22:46:48 -0500 (EST); g; /archives/html/s-news/2005-01/msg00164.html (8,798 bytes)
10. Re: really large files - import vs. use (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Thu, 16 Dec 2004 15:10:46 -0500 (EST); The read.table() function uses scan() to actually read the data (after getting the field headings, if any, from the first line). Note the following comment from the help file for scan(): As it reads; /archives/html/s-news/2004-12/msg00122.html (7,786 bytes)
11. Re: Making Graphs with clear printing fonts (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Thu, 4 Nov 2004 19:03:49 -0500 (EST); What Sundar describes is similar to my experience. With the driver installed, Word will import the EPS graphic but the visual quality is really poor. I suspect what is happening is that it converts i; /archives/html/s-news/2004-11/msg00026.html (10,195 bytes)
12. Re: Multiple R-Squared (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Sun, 4 Apr 2004 15:24:28 -0400 (EDT); The R-square is calculated based on variance explained relative to a null model. For a model with an intercept the null model includes an intercept (constant model), but for a model without an interc; /archives/html/s-news/2004-04/msg00008.html (9,696 bytes)
13. Re: scalability (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Mon, 29 Mar 2004 07:36:59 -0500 (EST); For this problem I wrote a little function many years ago called "dimsum" (which you can think of as summing over arbitrary dimensions, or as a Chinese breakfast, as you may prefer). "keep" are the; /archives/html/s-news/2004-03/msg00235.html (8,435 bytes)
14. *****SPAM***** Hi (score: 1): Author: zaslavsk@hcp.med.harvard.edu; Date: Tue, 10 Feb 2004 00:00:18 -0700; ??m?"?4Sc}?y?E?R???Yv"t1Th???h?E Ith?Ae??J-y?*?a`9?1?Q????,??d?$??>a$23y?0Ec$?e?I>????qI6?;?^?R??9Ugo8?>qO ????Z?~I&?N#?&z?U?FdR????y???^'???p?i#??V???????u???t???Ua?LyThOe/????6??dq??4?g?S???EkA?Eo; /archives/html/s-news/2004-02/msg00071.html (9,178 bytes)
15. s (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Thu, 15 Jan 2004 19:18:14 -0500 (EST); Power analysis tells you the probability that you would obtain a dataset that would give you evidence of a certain effect when the effect is of a hypothesized size. Why would you want to figure out t; /archives/html/s-news/2004-01/msg00082.html (8,373 bytes)
16. s (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Thu, 15 Jan 2004 20:22:03 -0500 (EST); Before an experiment, I might conjecture based on prior information that the standard error of my estimated effect will be 1, so the estimated effect would have to be about 2 to be significantly diff; /archives/html/s-news/2004-01/msg00084.html (9,061 bytes)
17. Re: sub-matrix and sub-data.frame (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Fri, 21 Nov 2003 21:26:56 -0500 (EST); g; /archives/html/s-news/2003-11/msg00142.html (8,372 bytes)
18. Mixing fonts/sizes/subscripts (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Wed, 1 Oct 2003 14:22:48 -0400 (EDT); I assume (perhaps incorrectly) the question is about putting these characters onto figures. Steve's suggestion is fairly platform-specific. Another alternative when preparing postscript plots is my; /archives/html/s-news/2003-10/msg00006.html (7,734 bytes)
19. Apparent spam/virus attack against S-news (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Tue, 19 Aug 2003 08:25:26 -0400 (EDT); Dear All, I've seen evidence that there is a spam/virus attack against addresses harvested off the S-news list. A number of familiar names from S-news (including my own) have been showing up in forge; /archives/html/s-news/2003-08/msg00101.html (7,241 bytes)
20. Re: Subscripting (score: 1): Author: Alan Zaslavsky <zaslavsk@hcp.med.harvard.edu>; Date: Fri, 6 Jun 2003 17:57:48 -0400 (EDT); For those who are curious, here is what is happening here: A floating point number is represented as a binary fraction with a fixed number of places and a 1 in the first place, times a power of 2. F; /archives/html/s-news/2003-06/msg00037.html (7,870 bytes)

Current List: 1 - 20
Page: [1] [2] [3] [4] [5] [6]