Allow for adjustment of precision/sig digits

[HeatherKates]

Hi! It would be great if there were an option for phyx programs to allow for user-specified prevision in output values (I'm thinking specifically of branch lengths). I have two issues that arise from the default - 1) The output trees are not always directly comparable with the input if it was generated from another program (this causes a problem when trying to search for subtrees, etc.), and 2) Very short branches are collapsed as polytomies. I know for that point the argument could be made that branches that short should be collapsed, but sometimes the bifurcation is needed for downstream analysis etc.

If there's already a way for a user to do this without altering the source code and I have missed it, please let me know.

Thank you!

[josephwb]

The significance is hard-coded; I believe we matched the RAxML code. I am also not thrilled by it (for the same reasons you state). I have been thinking of a way to keep/process the edge lengths as strings.
I will figure this out, but it is not a simple fix, and will take some time.