Re: [R] Rank-based p-value on large dataset

From: Sean Davis <sdavis2_at_mail.nih.gov>
Date: Fri 04 Mar 2005 - 09:49:55 EST

The x's and y's are different sets--210,000 values altogether. That is really the issue--they can't just be sorted, at least that I can see....

Sean

On Mar 3, 2005, at 5:38 PM, Huntsinger, Reid wrote:

> When you say the 130,000 points are from the empirical distribution,
> how did
> you get them? Is each one really one of the values of y? If you sorted
> y
> first, would you know which one (ie which index) each x is? (Sorting
> 80,000
> elements took essentially no time at all on my sub-gigahertz Pentium
> III.)
> But maybe that's not an option... more details would help.
>
> Reid Huntsinger
>
> -----Original Message-----
> From: r-help-bounces@stat.math.ethz.ch
> [mailto:r-help-bounces@stat.math.ethz.ch] On Behalf Of Sean Davis
> Sent: Thursday, March 03, 2005 5:22 PM
> To: r-help
> Subject: [R] Rank-based p-value on large dataset
>
>
> I have a fairly simple problem--I have about 80,000 values (call them
> y) that I am using as an empirical distribution and I want to find the
> p-value (never mind the multiple testing issues here, for the time
> being) of 130,000 points (call them x) from the empirical distribution.
> I typically do that (for one-sided test) something like
>
> loop over i in x
> p.val[i] = sum(y>x[i])/length(y)
>
> and repeat for all i. However, length(x) is large here as is
> length(y), so this process takes quite a long time. Any suggestions?

>
> Thanks,
> Sean
>
> ______________________________________________
> R-help@stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> http://www.R-project.org/posting-guide.html
>
>
>
>
>
> -----------------------------------------------------------------------
> -------
> Notice: This e-mail message, together with any attachment...{{dropped}}

R-help@stat.math.ethz.ch mailing list