R help - May 2014 - bootstrapping for maxium data

Dear friends,


I  have a numeric vector composed of 320 numbers.


Now, I want to do resample for 10000 times. I want to get maxium number for
every trial and get a 10000 maxium numbers.


I have tried to use "boot" package such as follows.


results<-boot(data=nearshore1,statistic=max,R=10000,stype="w")


But, surprising, the results shows that I got 10000 A (A is the maxium number in
"nearshore1")


It's a puzzle to me. I do not know how to do that. Thank you.


Yichun







	[[alternative HTML version deleted]]

Run the examples in the ?boot help page. Notice that boot needs a function that
accepts a dataframe and an index vector.

Sent from my iPhone
> On May 21, 2014, at 1:49 AM, ??? <yczhang at nigpas.ac.cn> wrote:
> 
> Dear friends,
> 
> 
> I  have a numeric vector composed of 320 numbers.
> 
> 
> Now, I want to do resample for 10000 times. I want to get maxium number for
every trial and get a 10000 maxium numbers.
> 
> 
> I have tried to use "boot" package such as follows.
> 
> 
> results<-boot(data=nearshore1,statistic=max,R=10000,stype="w")
> 
> 
> But, surprising, the results shows that I got 10000 A (A is the maxium
number in "nearshore1")
> 
> 
> It's a puzzle to me. I do not know how to do that. Thank you.
> 
> 
> Yichun
> 
> 
> 
> 
> 
> 
> 
>    [[alternative HTML version deleted]]
> 
> ______________________________________________
> 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.

Yichun,

1. You are doing this wrong!
2. It is the wrong thing to do!

In more detail:
1. To use boot, it helps to read the help file which clearly says the 
statistic must be a function of two arguments, the original dataset and 
the indices saying which of the original datapoints are in the bootstrap 
dataset (or the frequency of each original datapoint in the bootstrap 
dataset). A valid statistic would be
function(x,i) max(x[i])

2. See Davison and Hinkley example 2.15 which shows that the bootstrap 
is inconsistent for the sample maximum!


On 21/05/2014 1:49 AM, ??? wrote:
> Dear friends,
>
>
> I  have a numeric vector composed of 320 numbers.
>
>
> Now, I want to do resample for 10000 times. I want to get maxium number for
every trial and get a 10000 maxium numbers.
>
>
> I have tried to use "boot" package such as follows.
>
>
> results<-boot(data=nearshore1,statistic=max,R=10000,stype="w")
>
>
> But, surprising, the results shows that I got 10000 A (A is the maxium
number in "nearshore1")
>
>
> It's a puzzle to me. I do not know how to do that. Thank you.
>
>
> Yichun
>
>
>
>
>
>
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> 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.

-- 
------------------------------------------------------------------
|   Angelo J. Canty                Email: cantya at mcmaster.ca     |
|   Mathematics and Statistics     Phone: (905) 525-9140 x 27079 |
|   McMaster University            Fax  : (905) 522-0935         |
|   1280 Main St. W.                                             |
|   Hamilton ON L8S 4K1                                          |