Skip to content

bootstrap question

2 messages · Erin Hodgess, Roger D. Peng

Erin Hodgess
# 
Dear r People


I have a bootstrap question, please.

(this may possibly be a problem with my understanding of the bootstrap)

Suppose I have a sample of size 15, x[1], ...x[15].

This is a sample which is small compared to the population.

by the way, the x[i] are iid and have a common distribution F.

As I understand the bootstrap, that function will take a user specified
number of repititions.  In each repitition, each element x[i] has an
equal chance of selection.

Suppose I want to estimate the mean.

In each rep, a new sample is generated.  the mean is calculated in each rep.


I would like to see those means.

If I use the boot command,

boot(x,mean,R=40)$t

I thought that I would obtain 40 different values.  However, those $t values
are all the same.

what am I doing wrong, please?

thanks yet again!

sincerely,
Erin Hodgess
mailto: hodgess at uhddx01.dt.uh.edu
Roger D. Peng
# 
You are using the boot function incorrectly.  This is taken from the 
help page:


statistic: A function which when applied to data returns a vector
           containing the statistic(s) of interest.  When
           `sim="parametric"', the first argument to `statistic' must be
           the data.  For each replicate a simulated dataset returned by
           `ran.gen' will be passed.  In all other cases `statistic'
           must take at least two arguments.  The first argument passed
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
           will always be the original data. The second will be a vector
           of indices, frequencies or weights which define the bootstrap
           sample.

The default is nonparametric bootstrap (sim = "ordinary"), so you need 
to specify a function with TWO arguments.  Try

x <- rnorm(100)
boot(x, function(x, i) mean(x[i]), R = 1000)$t

-roger
Erin Hodgess wrote: