roimethstat and order of chromosomes

[andrewdavidsmith]

It seems roimethstat has problems when the chromosomes are not in the same order between the methylation file and the regions file.

[andrewdavidsmith]

This likely arises because the entire data set is processed as a single vector with all chromosomes together. Separating the chroms, as long as they are in correspondence when processed, will circumvent this problem.

[andrewdavidsmith]

Candidate solution: modify the regions_bounds function so that it takes iterators, or at least a start and end index for the range. Then, when loading the methylation data, keep some index for the start and end of each chromosome's sites within the vector. Before looking for the appropriate bounds for any given region, first make sure to be searching within the appropriate bounds among the sites. Then the order of chroms will not matter, but the chroms must still be together and the sites must still be sorted (or all bets are off).