[R] more efficient double summation...

[Charles C. Berry]

On Fri, 29 Aug 2008, Giovanni Petris wrote:

>
> You can cut execution time by a factor 2 simply using the fact that the
> double summation is symmetric in the indices j and k:
>
> 2 * sum(sapply(1:(m-1), function(k){sum(sapply((k-1):m,
>  function(j){x[k]*x[j]*dnorm((mu[j]+mu[k])/sqrt(sig[k]+sig[j]))/sqrt(sig[k]+sig[j])}))}))
>  + sum(x^2*dnorm((2*mu)/sqrt(2*sig))/ sqrt(2*sig))
>
> Best,
> Giovanni Petris
>

Better still, vectorize kathie's code as:


 	foo <- function(j,k) {x[k]*x[j]*dnorm((mu[j]+mu[k])/
 		sqrt(sig[k]+sig[j]))/sqrt(sig[k]+sig[j])}

 	sum( outer(1:m, 1:m, foo ) )

To achieve Giovanni's savings in a vectorized framework use

 	jk <- combn( 1:m, 2 )
 	j.alone <- 1:m
 	2*sum( foo( jk[1,], jk[2,]) )+ sum( foo( j.alone, j.alone ) )

HTH,

Chuck

>> Date: Thu, 28 Aug 2008 21:47:38 -0700 (PDT)
>> From: kathie <kathryn.lord2000 at gmail.com>
>> Sender: r-help-bounces at r-project.org
>> Precedence: list
>>
>>
>> Dear R users...
>>
>> I made the R-code for this double summation computation
>>
>>
>> http://www.nabble.com/file/p19213599/doublesum.jpg
>>
>> -------------------------------------------------
>> Here is my code..
>>
>>
>> sum(sapply(1:m, function(k){sum(sapply(1:m,
>> function(j){x[k]*x[j]*dnorm((mu[j]+mu[k])/sqrt(sig[k]+sig[j]))/sqrt(sig[k]+sig[j])}))}))
>>
>>
>> -------------------------------------------------
>>
>> In fact, this is a part of optimization. I think if it is changed more
>> efficiently, then the running time could be shortened.
>>
>> How could I change this to more efficiently? Any suggestion will be greatly
>> appreciated.
>>
>> Kathryn Lord
>> --
>> View this message in context: http://www.nabble.com/more-efficient-double-summation...-tp19213599p19213599.html
>> Sent from the R help mailing list archive at Nabble.com.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>>
>
> -- 
>
> Giovanni Petris  <GPetris at uark.edu>
> Associate Professor
> Department of Mathematical Sciences
> University of Arkansas - Fayetteville, AR 72701
> Ph: (479) 575-6324, 575-8630 (fax)
> http://definetti.uark.edu/~gpetris/
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

Charles C. Berry                            (858) 534-2098
                                             Dept of Family/Preventive Medicine
E mailto:cberry at tajo.ucsd.edu	            UC San Diego
http://famprevmed.ucsd.edu/faculty/cberry/  La Jolla, San Diego 92093-0901