Markus Werle wrote:
Many thanks for this. After a first glance: This looks like a lib that is useful for me immediately. Again OpenSource boosts (sic!) productivity.
Good :-)
Q1: In the docs you mention the accumulator package from Eric Niebler. Is this simply a complement package or is there some overlap, which might be merged _before_ inclusion into boost? I simply fear the same difficulties as for boost::(*|lambda)::bind, the placeholder chaos (*|boost::*)::_1, and all the joy to encounter with ADL.
No overlap, Erics package computes mean, sd, variance etc in a highly optimised manner. Once you have those stats you can feed them into the toolkit to do your statistical tests. Or you can compute those stats using whatever other method you currently use. The two packages should be very complementary.
Q2: if the docs contain a "TODO" or "Under construction" marker does that mean the code still has to written or does it only mean the docs still have to be written?
It means I've been lazy in writing the docs :-) The main blank spaces were in the tutorial/examples I think ? If you have good suggestions for examples to go in there that would be great. Or was it something else you had in mind? Thanks for the feedback, John.