Hi.
It could be considered strange that CM reports 0.2 (instead of some value
close to 1) as the 10-percentile of the { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 }
array.
Quoting the Javadoc:
---
For large samples, the different methods
agree closely, but when sample sizes are small, different methods will
give significantly different results.
---
Nevertheless, shouldn't such a difference from the expected result lead to
wonder whether the implementation should switch to another algorithm for
small datasets?
Thanks,
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
