:tocdepth: 3

.. _guppy_compress:

========
compress
========

`compress` compresses a placefile's pqueries.

::

  usage: compress [options] placefile

Options
=======

--point-mass  Treat every pquery as a point mass concentrated on the highest-weight placement.
--pp  Use posterior probability for the weight.
-o  Specify the filename to write to.
--out-dir  Specify the directory to write files to.
--prefix  Specify a string to be prepended to filenames.
--cutoff  The cutoff parameter for mass compression
--discard-below  In island clustering, ignore pquery locations with a mass less than the specified value.
--mcl  Use MCL clustering instead of island clustering.
--inflation  If specified, pass this as the inflation value to MCL.

Details
=======

A cutoff *c* is specified via a command line flag.
The `compress` command merges pairs of pqueries that have KR distance between them less than *c*.

To compress the pqueries:

* divide the pqueries into sets via :doc:`guppy_islands` or :doc:`guppy_mcl`

* for each pquery set, calculate all of the pairwise distances between the pqueries for that island and put them in a matrix
* build a graph such that the nodes are pqueries, and there is an edge between them if their distance is less than *c*
* merge pqueries according to this graph as described below.


Each original pquery (=node) will get merged into one of the the selected pqueries.
This will happen as follows.
Maintain a set of unmerged pqueries, and a set of pairs *(w, d(w))*, where *w* is a selected pquery and *d(w)* is the degree of *w* in the graph.

* find the *(w, d(w))* pair with the greatest d(w) and remove it from the set
* find all of the unmerged pqueries that are adjacent to *w* in the graph, and merge their mass into *w*. Remove *w* and all of the adjacent pqueries from the unmerged pquery set.
* repeat!

Stop when the unmerged pquery set is empty.