Is there a library that implements a fast maximum rank correlation estimator?

I'm looking for an implementation of a fast maximum rank correlation (MRC) estimator. This will be applied to large-ish sparse matrices (~100,000 by 10,000) in a text-mining application.

I'm working in python and R, so it would be nice to find something in those languages. Failing th开发者_StackOverflow社区at, I could probably convert code from some other language.

Any suggestions?

The answer is no. Having read up on the subject, MRC estimators are famously hard to compute, although there is hope that maybe better algorithms could be designed.

I'll post this answer to CV as well.