qDiptab {diptest} | R Documentation |
Whereas Hartigan(1985) published a table of empirical percentage
points of the dip statistic (see dip
) based on N=9999
samples of size n from U[0,1], our table of empirical
quantiles is currently based on N=100'001 samples for each n.
A numeric matrix where each row corresponds to sample size n, and each column to a probability (percentage) in [0,1]. The dimnames are coercable to these values, see the examples.
Taking N=100'001 ensures that all the quantile(X, p)
used here are exactly order statistics sort(X)[k]
.
Martin Maechler maechler@stat.math.ethz.ch
dip
, also for the references.
data(qDiptab) str(qDiptab) ## the sample sizes `n' : (nn <- as.integer(rownames(qDiptab))) ## the probabilities: P.p <- as.numeric(print(colnames(qDiptab))) ## This is as "Table 1" in Hartigan & Hartigan (1985) -- but more accurate ps <- c(1,5,10,50,90,95,99, 99.5, 99.9)/100 tab1 <- qDiptab[nn <= 200, as.character(ps)] round(tab1, 4)