VV_VVm: Variances of the Expected Frequency Spectrum (zipfR)
In zipfR: Statistical Models for Word Frequency Distributions

Description Usage Arguments Details Value See Also Examples

VV and VVm are generic methods that can (and should) be used to compute the variance of the vocabulary size and the variances of spectrum elements according to an LNRE model (i.e. an object of class lnre). These methods are also used to access variance information stored in some objects of class spc and vgc.

1 2	VV(obj, N=NA, ...) VVm(obj, m, N=NA, ...)

`obj`	an object of class `lnre` (LNRE model), `spc` (frequency spectrum) or `vgc` (vocabulary growth curve).
`m`	positive integer value determining the frequency class m for which variances are returned (or a vector of such values).
`N`	sample size N for which variances are calculated (`lnre` objects only)
`...`	additional arguments passed on to the method implementation (see respective manpages for details)

spc and vgc objects must represent an expected or interpolated frequency spectrum or VGC, and must include variance data.

For vgc objects, the VVm method allows only a single value m to be specified.

The argument N is only allowed for LNRE models and will trigger an error message otherwise.

For a LNRE model (class lnre), VV computes the variance of the random variable V(N) (vocabulary size), and VVm computes the variance of the random variables V_m(N) (spectrum elements), for a sample of specified size N.

For an observed or interpolated frequency spectrum (class spc), VV returns the variance of the expected vocabulary size, and VVm returns variances of the spectrum elements. These methods are only applicable if the spc object includes variance information.

For an expected or interpolated vocabulary growth curve (class vgc), VV returns the variance vector of the expected vocabulary sizes V, and VVm the corresponding vector for V_m. These methods are only applicable if the vgc object includes variance information.

For details on the implementations of these methods, see VV.spc, VV.vgc, etc.

Expected vocabulary size and frequency spectrum for a sample of size N according to a LNRE model can be computed with the analogous methods EV and EVm. For spc and vgc objects, V and V_m are always accessed with the methods V and Vm, even if they represent expected or interpolated values.