myresamplefn {cabootcrs} | R Documentation |
Example of a user-generated resampling routine.
Description
myresamplefn
in this case assumes that each pair of cells represents 50 people
answering yes or no to a question, with undecideds not recorded
Usage
myresamplefn(X)
Arguments
X |
a data matrix, to be resampled from |
Details
This is only intended as an example of a user-generated resampling routine, users should replace it with their own function of this name
In this example we assume that rows groups of 50 people have each been asked columns/2 questions, with possible answers yes/no/undecided. It uses binomial bootstrapping for pairs of columns, assuming that successive columns are "yes" and "no" answers, with others undecided, with sums of pairs of columns having a maximum, in this case 50.
Value
a resampled version of the input data matrix
See Also
Examples
# Five groups of people answer two yes/no/undecided questions
# Note: this is just an example, and does not intend to claim that this
# is the correct analysis for such a data set
x <- as.matrix( rbind( c(22,25,18,22), c(12,23,21,27),
c(31,12,28,22), c(29,14,35,11), c(7,31,12,21)))
xresampled <- myresamplefn(x)
bmr <- cabootcrs(x,"myresample",nboots=199)
[Package cabootcrs version 2.1.0 Index]