R help - Nov 2013 - Generating bootstrap samples from a panel data frame

With a data frame (call it *d*) composed of 2000 individuals and
*n*observations for each individual (thus
*2000n* observations in total), I would like to generate *k* bootstrap
samples with replacement from *d*. Amongst other variables, *d* has a
numeric variable *id* taking on identical value for observations belonging
to the same individual.

Taking into consideration the panel nature of the data, I want to generate
many bootstrap samples with replacement and store each bootstrap sample
data frame for further use. Sampling (or selection into the bootstrap
sample) shall be based on individuals (on unique values of *id*) such that
if an individual is in a particular bootstrap sample, so will all
observations belonging to that individual.

How can I do this in r?

	[[alternative HTML version deleted]]

With a sample data frame:
d = data.frame(id = rep(c(1,2,3,4),5), x = round(rexp(20), digits=2), y
round(runif(20), digits=2))

I would like to generate 100 bootstrap data with replacement and save each
bootstrap data as b1, b2, ..., b100. I attempted the sample function but
could not make it work



On Mon, Nov 11, 2013 at 6:11 PM, Dereje Fentie <derex.fen@gmail.com>
wrote:
>
>   With a data frame (call it *d*) composed of 2000 individuals and
*n*observations for each individual (thus
> *2000n* observations in total), I would like to generate *k* bootstrap
> samples with replacement from *d*. Amongst other variables, *d* has a
> numeric variable *id* taking on identical value for observations
> belonging to the same individual.
>
> Taking into consideration the panel nature of the data, I want to generate
> many bootstrap samples with replacement and store each bootstrap sample
> data frame for further use. Sampling (or selection into the bootstrap
> sample) shall be based on individuals (on unique values of *id*) such
> that if an individual is in a particular bootstrap sample, so will all
> observations belonging to that individual.
>
> How can I do this in r?
>
	[[alternative HTML version deleted]]