Package wish list

Add the possibility to use a multiPhylo object in time.series.
A jackknife.ordination function for jackknifing the pre-ordination matrix and feed an ordination pipeline (paralel-able and out-of-R-able)
Add mvtnorm and Pie and Weitz's 2001 BRW null models.
From Anderson et al 2012 Nature: implement the following in Claddis? :"strength of association between the the 11 coordinate axes (CA) of the NMDS and the biomechanical characters using Spearman rank-order (for continuous characters) and Mann- Whitney U (for the discrete characters) tests".
Add correlation with sample size for disparity metrics (calculate Spearman’s rho for the correlation between the metric and the sample size)
Test from Anderson: We applied paired-sample t tests as outlined in Zelditch et al. (2004: 222)
We also calculated marginal likelihoods for variance as an alternative test for differences in functional disparity between stages. where L is the likelihood, n is sample size, σ2 is the estimated variance parameter, and s2 is the variance
Changes in disparity between stage bins follows the procedure outlined in Finarelli and Flynn (2007).

To do!

Improve test.dispRity architecture
Change the structure of the dispRity calculation loop to be run in C. make.metric would then be a compiler function.
In manual, describe dataset + deal with comments