Bootstrap can be relatively slow

[lenthomas]

Bootstrapping a large dataset is relatively slow. Compare the bootstrap execution time in the CTDS example with that of a similar analysis of the DistWin duiker example dataset -- the latter is very much faster.

There are likely overheads involved in setting up and executing the bootstrap in the R code (which is in mrds) that could be optimized to make it faster -- although this possibly will come at the cost of generality/robustness.

I'm marking this as a low priority enhancement -- something possibly to look at in the future.

More potentially useful information is in issue 168.