Comments for BayesRules http://bayesrules.wordpress.com Just another WordPress.com weblog Tue, 17 Nov 2009 18:38:58 +0000 http://wordpress.com/ hourly 1 Comment on HCOL 195 11/09/09 by I’ve commented on Monday’s class « BayesRules http://bayesrules.wordpress.com/2009/11/11/hcol-195-110909/#comment-76 I’ve commented on Monday’s class « BayesRules Tue, 17 Nov 2009 18:38:58 +0000 http://bayesrules.wordpress.com/?p=727#comment-76 [...] commented on Monday’s class By bayesrules I’ve commented on Monday’s class here. Possibly related posts: (automatically generated)Darkness [...] [...] commented on Monday’s class By bayesrules I’ve commented on Monday’s class here. Possibly related posts: (automatically generated)Darkness [...]

]]> Comment on URLs mentioned on Thursday by Jeff http://bayesrules.wordpress.com/2009/04/17/urls-mentioned-on-thursday/#comment-24 Jeff Fri, 17 Apr 2009 17:51:53 +0000 http://bayesrules.wordpress.com/?p=520#comment-24 correction: comments start on page 335. correction: comments start on page 335.

]]> Comment on URLs mentioned on Thursday by Jeff http://bayesrules.wordpress.com/2009/04/17/urls-mentioned-on-thursday/#comment-23 Jeff Fri, 17 Apr 2009 17:51:30 +0000 http://bayesrules.wordpress.com/?p=520#comment-23 for paper listed in item #1, it is also worthwhile to read the comments on the article by several prominent statisticians, including d.r. cox. the comments start on page 337. for paper listed in item #1, it is also worthwhile to read the comments on the article by several prominent statisticians, including d.r. cox. the comments start on page 337.

]]> Comment on Stat 295, April 16, 2009 by bayesrules http://bayesrules.wordpress.com/2009/04/17/stat-295-april-16-2009/#comment-22 bayesrules Fri, 17 Apr 2009 17:49:40 +0000 http://bayesrules.wordpress.com/?p=514#comment-22 I take Jeff's point. Please note that <a href="http://www.jstor.org/stable/2245772" rel="nofollow">Berger and Delampady</a> have an extensive discussion of this (and other) points in their paper. See Section 4.5. I take Jeff’s point.

Please note that Berger and Delampady have an extensive discussion of this (and other) points in their paper. See Section 4.5.

]]> Comment on STAT 295 4/7/09 by bayesrules http://bayesrules.wordpress.com/2009/04/08/stat-295-4709/#comment-17 bayesrules Thu, 09 Apr 2009 14:45:40 +0000 http://bayesrules.wordpress.com/?p=445#comment-17 This is an interesting article, worth reading. Bill This is an interesting article, worth reading.

Bill

]]> Comment on STAT 295 4/7/09 by Jeff http://bayesrules.wordpress.com/2009/04/08/stat-295-4709/#comment-14 Jeff Wed, 08 Apr 2009 19:26:23 +0000 http://bayesrules.wordpress.com/?p=445#comment-14 bill and everyone in class, regarding the questions about model assessment bill is raising--i think the paper by r little in the american statistician, 2006 number 3, gives a good discussion of how a bayesian might assess his/her models. if you have a look, let me know what you think. jeff bill and everyone in class,

regarding the questions about model assessment bill is raising–i think the paper by r little in the american statistician, 2006 number 3, gives a good discussion of how a bayesian might assess his/her models.

if you have a look, let me know what you think.

jeff

]]> Comment on Short assignment, due 3/26 by Stat 295, 3/24/09 « BayesRules http://bayesrules.wordpress.com/2009/03/24/short-assignment-due-326/#comment-9 Stat 295, 3/24/09 « BayesRules Wed, 25 Mar 2009 20:03:56 +0000 http://bayesrules.wordpress.com/?p=318#comment-9 [...] BayesRules Just another WordPress.com weblog « Short assignment, due 3/26 [...] [...] BayesRules Just another WordPress.com weblog « Short assignment, due 3/26 [...]

]]> Comment on Comment Problem Set #4, problem 3 by bayesrules http://bayesrules.wordpress.com/2009/02/23/comment-on-the-third-problem/#comment-5 bayesrules Fri, 06 Mar 2009 19:56:46 +0000 http://bayesrules.wordpress.com/?p=98#comment-5 I'll outline the procedure. The likelihood is $latex (4\lambda)^{y_i} e^{-4\lambda}$ where $latex y_i$ is the number of units sold in geographic area $latex i$. The priors for the two areas are different, but are gammas with (for example in area 1) parameters 144 and 2.4, so in this instance $latex p(\lambda_1)=\lambda_1^{144-1}e^{-2.4\lambda_1}$. Multiplying prior times likelihood we get another gamma, this time with parameters $latex 144+y_1$ and $latex 2.4+4$. So to get a sample on $latex \lambda_1$ simply use the R function rgamma with these parameters to generate a large sample $latex \{\lambda_1^j\}, j=1,...,N$ Similarly generate a sample of $latex \lambda_2^j$'s of the same size. Subtract 1.5 times the samples of the $latex \lambda_2$'s from the samples of the $latex \lambda_1$'s to get a sample of the differences. To do this, subtract 1.5 times the first component of sample #2 from the first component of sample #1, and so forth, to get a sample of differences, again of length N. This can be done by simply subtracting 1.5 times the second vector from the first. Then in R, the proportion of the difference that is greater than 0 is the answer you want. I’ll outline the procedure. The likelihood is (4\lambda)^{y_i} e^{-4\lambda} where y_i is the number of units sold in geographic area i. The priors for the two areas are different, but are gammas with (for example in area 1) parameters 144 and 2.4, so in this instance p(\lambda_1)=\lambda_1^{144-1}e^{-2.4\lambda_1}. Multiplying prior times likelihood we get another gamma, this time with parameters 144+y_1 and 2.4+4. So to get a sample on \lambda_1 simply use the R function rgamma with these parameters to generate a large sample \{\lambda_1^j\}, j=1,...,N

Similarly generate a sample of \lambda_2^j’s of the same size.

Subtract 1.5 times the samples of the \lambda_2’s from the samples of the \lambda_1’s to get a sample of the differences. To do this, subtract 1.5 times the first component of sample #2 from the first component of sample #1, and so forth, to get a sample of differences, again of length N. This can be done by simply subtracting 1.5 times the second vector from the first. Then in R, the proportion of the difference that is greater than 0 is the answer you want.

]]> Comment on Comment Problem Set #4, problem 3 by xyz http://bayesrules.wordpress.com/2009/02/23/comment-on-the-third-problem/#comment-4 xyz Fri, 06 Mar 2009 14:10:06 +0000 http://bayesrules.wordpress.com/?p=98#comment-4 Can you post solution of this example?? Can you post solution of this example??

]]>