Compute Sample variance for dependent variable. For repeated response the sample variance of predictor compute over a cross section of predictor, for non repeated data sample variance of response computed by difference of residuals of consequetive values.
Usage
zvalues(res, ni, xo)
Arguments
res
The data to compute variance for. In most of application it is residuals of fitt.
ni
vector of bumber of repeated data. It can be output from nonrepl function.
xo
Position of the repeated data in original vector.
Value
vector of classic variance.
Details
Typically it is used to compute the variance of residuals output from nonrepl function.
References
Bunke, O., Droge, B., Polzehl, J. Splus tools for model selection in nonlinear regression (1998) Computational Statistics, 13 (2), pp. 257-281.