let's say I have a 3 float32 matrix w/ 5000 columns and 1 row.
I've been searching around for, but so far haven't found a way to:
1. delete duplicate cells/columns.
2. delete nearly duplicate cells/columns based on a similarity threshold. i.e. if the plane 0 & 1 value differences are < .1, of another cell's values.
possible? I have a feeling some of Wesley's xray objects may come into play here...
