"best" parameter in the get.placements method for tree

[Jacob Nearing]

Hello everyone,

First of all thank you for making this open source software it has been extremely useful in our research. 

My question is why in the function get.placements the "best" parameter seems to filter placements based on the minimum log likelihood. Should it not filter based on the maximum log likelihood to identify the best/most likely placements (as the pplacer option --keep-at-most X does)?

Also if the get.placements function could be updated to describe what the "best" parameter filters on that would be extremely useful (had to look through source code to identify this).

Cheers,

Jacob