[BioC] Simulation of the beta average

[Shi, Tao]

Why can you generate simulation data by bootstrapping your existing data?

...Tao




----- Original Message ----
> From: Samor Gandhi <samorgandhi at yahoo.com>
> To: Sean Davis <sdavis2 at mail.nih.gov>
> Cc: "bioconductor at stat.math.ethz.ch" <bioconductor at stat.math.ethz.ch>
> Sent: Tue, September 28, 2010 6:23:56 AM
> Subject: Re: [BioC] Simulation of the beta average
> 
> Hi Sean,
> 
> Thank you for your reply. What I posted would be for one sample,  but I need to 
>have let say 20 samples into two groups, that the first group  would have 
>different mean and standard deviation than the second group; and that  was my 
>problem how to generate these.
> 
> I did something like 
> 
> S  <- 10000
>  n <- 10
>  data1 <- c(runif(n*S*0.2,0,0.05),  
>runif(n*S*0.6,0.06,0.95),runif(n*S*0.2,0.96,1))
>  data2 <-  c(runif(n*S*0.2,0.06,0.1), 
>runif(n*S*0.6,0.11,0.8),runif(n*S*0.2,0.81,0.9))
> 
>  group1 <- matrix(data1, nrow = S,ncol = n)
>  group2 <-  matrix(data2, nrow = S,ncol = n)
> 
>  wt <- rep(NA, S)
>  for (i in 1:S)  {
>  wt[i] <- wilcox.test(group1[i, ], group2[i, ], alternative =  
>"two.sided")$p.value }
>  Wsize <- sum(wt < 0.05)/S
>  Wsize
> 
> I  very thankful for any help.
> 
> Samor
> 
> --- On Tue, 28/9/10, Sean Davis  <sdavis2 at mail.nih.gov>  wrote:
> 
> From: Sean Davis <sdavis2 at mail.nih.gov>
> Subject: Re:  [BioC] Simulation of the beta average
> To: "Samor Gandhi" <samorgandhi at yahoo.com>
> Cc: "bioconductor at stat.math.ethz.ch"  <bioconductor at stat.math.ethz.ch>
> Date:  Tuesday, 28 September, 2010, 17:46
> 
> 
> 
> On Tue, Sep 28, 2010 at 7:15  AM, Samor Gandhi <samorgandhi at yahoo.com>  wrote:
> 
> Hi Sean,
> 
> I am trying to compute the power of the SAMr for  t-test and Wilcoxon test!
> 
> 
> 
> Hi, Samor.
> You might need to look  into the data a bit more and perhaps review a bit about 
>power calculation before  proceeding further.  There are a few steps missing, I 
>think.  What you posted  doesn't constitute a simulation that could be used for 
>power calculations unless  I misunderstand what you are trying to  do.
> 
> Sean
>  
> Regards,
> Samor
> 
> --- On Tue, 28/9/10, Sean Davis  <sdavis2 at mail.nih.gov>  wrote:
> 
> 
> From: Sean Davis <sdavis2 at mail.nih.gov>
> Subject: Re:  [BioC] Simulation of the beta average
> To: "Samor Gandhi" <samorgandhi at yahoo.com>
> 
> Cc: bioconductor at stat.math.ethz.ch
> Date:  Tuesday, 28 September, 2010, 14:31
> 
> 
> 
> On Tue, Sep 28, 2010 at 4:45  AM, Samor Gandhi <samorgandhi at yahoo.com>  wrote:
> 
> 
> Hello,
> 
> 
> 
> I would like to run a simulation study  on the beta average of the Beta Studio 
>illumin. I did something like x <-  c(runif(200,0,0.1), 
>runif(600,0.11,0.89),runif(200,0.9,1)); hist(x)! Would this  be a reasonable 
>simulation?
> 
> 
> 
> 
> Hi, Samor.  
> I think the  answer to your question is going to depend on what you mean by 
>"reasonable" and  that will depend on the question you are actually trying to 
>answer.  What is the  question you are trying to answer?  Are you trying to 
>perform hypothesis testing  or algorithm development, or something else?   
> 
> 
> Sean 
> 
> 
> 
> 
> 
>       
> 
> 
> 
> 
>       
>     [[alternative  HTML version deleted]]
> 
>