Calculate Jackknife Statistics

DESCRIPTION:

Utility function called by jackknife to calculate jackknife statistics and create an object of class jackknife which inherits from resamp.

USAGE:

jackstats(replicates, observed, n, call, seed.start, dim.obs=NULL) 

REQUIRED ARGUMENTS:

replicates
matrix of jackknife replicates of the parameters of interest, with one column per parameter.
observed
parameter values for the original set of data.
n
number of observations in the original set of data, or number of groups if the data is jackknifed with group.size not equal to one.
call
call used by function which calls jackstats.
seed.start
random seed set at start of function calling jackstats. Because the resampling indices are not random with jackknifing, this seed is only of interest when the statistic uses randomization, or group.size is not equal to one.

OPTIONAL ARGUMENTS:

dim.obs
if the statistic of interest returns a matrix of parameters, the dimensions of this matrix are stored as dim.obs. Used in plot.resamp and qqnorm.resamp to control the layout of plots.

VALUE:

an object of class jackknife which inherits from resamp. This has components call, observed, replicates, estimate, B, n, dim.obs, and seed.start. The data frame estimate has three columns containing the bootstrap estimates of Bias, Mean, and SE.

REFERENCES:

Efron, B. and Tibshirani, R. J. (1993). An Introduction to the Bootstrap. San Francisco: Chapman & Hall.

Shao, J. and Tu, D. (1995). The Jackknife and Bootstrap. New York: Springer-Verlag.

SEE ALSO:

.