bootstrapT:
Calculate bootstrap p-values for t statistics
Description
This function takes a numerical matrix and column indexes for two
groups to calculate bootstrapped (by re-sampling) p-values comparing the
equality of means from the two groups.
Usage
bootstrapT(x, k=20000, obs1, obs2, ...)
Arguments
x
numerical matrix to be bootstrapped. The t statistics is
calculated by row using the column indexes given by obs1 and
obs2 for the two groups tested.
k
number of bootstrap re-samplings to be done. Defaults to 20000.
obs1
logical or numerical column indexes of the first group.
obs2
logical or numerical column indexes of the second group.
...
additional parameters for t.test
function from package stats.
Value
The result of this function is a numerical matrix with number of rows
given by the rows of the argument x and 3 columns. The first
column contain the difference of means between the two groups, the
second one contain the original t statistic and the last one gives the
bootstrapped p-values, for all rows of the matrix x.