Does anyone have any opinions about the general reliability/usability of the mathematical functions in the boost library, for example the inverse of the cdf for t distributions and the inverse incomplete beta function etc?
How does it compare with commercially available products such as the matlab toolboxes?
[The reason I'm asking is that I'm interested in hearing such opinions. So please don't reply to this thread by saying something like "The answer to your question is yes, some people do have opinions about the boost library. However, I am not such a person."]