Skip to content
Prev 181077 / 398506 Next

minmun p-value for wilcox.test and correlation

Peter Flom wrote:
I'm also curious. Anyway, you can get the calculated value by:

wilcox.test(....)$p.value

Note that printing of htest objects results always in the above 
mentioned minimum for very small values, because we cannot be sure about 
numerical precision, particularly also in the far from centre regions of 
a distribution... (and since you should not report a p-value of exactly 
0 that may numerically arise when doing such computations).

Uwe Ligges