compareBoot {ssExtra} | R Documentation |
This routine will run testBoot
on each of n+B
combinations. Thus, it can be used to look at the differences between
the two bca confidence interval routines,
bcaboot:bcajack
and
boot:boot.ci
for confidence interval calculation.
compareBoot(n = c(100, 200), B = c(1000, 1200), tablePath = "", tableName = "", digits = c(0, 0, 0, rep(3, 6)), caption = "Bootstrap Confidence Interval Comparison", label = "tab:bootTable", ssshhh = FALSE, ...)
n |
A vector of sample sizes that are to be used for
bootstrapping. Note that since this routine calls
|
B |
A vector with the number of bootstrap sample replicates at each sample size. |
tablePath |
The relative path to |
tableName |
The file name for the xtable output table. Use
‘’ for no hardcopy table output; otherwise a |
digits |
The number of display digits for the
|
caption |
The caption for the |
label |
The LaTeX label for the |
ssshhh |
|
... |
Arguments passed to |
This routine was written simply to facilitate comparison the two bca
confidence interval results in the different packages cited above. Note
that the results can be saved to a file by using the
xtable
function in package xtable.
A list with...
df |
A data frame with the results, which should be self-explanatory. |
xtab |
A LaTeX table object of class “xtable”. |
Neither this routine, nor testBoot
are meant to be general
purpose routines. They were written to explore the differences in
calculated bca intervals for a specific population, that of a standard
normal. One could certainly adapt these routines to other uses.
Please see the note in testBoot
concerning the potential
error that can occur using boot.ci
.
Jeffrey H. Gove
testBoot
,
bcaboot
,
bcajack
,
boot
,
boot.ci
# # simply... # ## Not run: cmp.bs = compareBoot() ## End(Not run)