| sample.spc {zipfR} | R Documentation |
Compute incremental random samples from a frequency spectrum (an object
of class spc).
sample.spc(obj, N, force.list=FALSE)
obj |
an object of class spc, representing a frequency
spectrum |
N |
a vector of non-negative integers in increasing order, the sample sizes for which incremental samples will be generated |
force.list |
if TRUE, the return value will always be a
list of spc objects, even if N is just a single
integer |
This function is currently implemented as a wrapper around
sample.tfl, using spc2tfl and tfl2spc to convert
between frequency spectra and type frequency lists. A direct
implementation might be slightly more efficient, but would very likely
not make a substantial difference.
If N is a single integer (and the force.list flag is not
set), a spc object representing the frequency spectrum of a
random sample of size N from obj.
If N is a vector of length greater one, or if
force.list=TRUE, a list of spc objects representing the
frequency spectra of incremental random samples of the specified sizes
N. Incremental means that each sample is a superset of
the preceding sample.
spc for more information about frequency spectra
sample.tfl is an analogous function for type frequency
lists (objects of class tfl)
sample.spc takes a single concrete random
subsample from a spectrum and returns the spectrum of the subsample,
unlike spc.interp, that computes the expected
frequency spectrum for random subsamples of size N by
binomial interpolation.
## read Brown spectrum data(Brown.spc) summary(Brown.spc) ## sample a spectrum of 100k tokens MiniBrown.spc <- sample.spc(Brown.spc,1e+5) summary(MiniBrown.spc) ## if we repat, we get a different sample MiniBrown.spc <- sample.spc(Brown.spc,1e+5) summary(MiniBrown.spc)