storing packing results in memory

cellPack results data structures BEFORE

env
    .molecules  # array of [(x,y,z), rotation, ingredient, radius]
    .result  # array of [numpy.array(jtrans).flatten().tolist(), numpy.array(rotMatj),ingredient,ptInd]

## used for search tree lookup, all ingredients regardless of container :
    .rTrans # array of positions 
    .rRot # array of rotations
    .rIngr # array of ingredients 

containers 
    .molecules # array of [(x,y,z), rotation, ingredient, radius]

ingredient
    .results # an array of [position, rotation]

Questions:

• is it important to store results on the containers?
• do we need to actually store the whole ingredient object each time? could we just store the id?
• what's the best structure? the group of flat arrays (rTrans, rRot etc), or one array of objects [{p: , r:, id: pt_id: etc}] or array or arrays as we have in .result?
• Should we store everything as np arrays?

TODO:

• we should have one "store" function that stores the data (right now each place function updates .rTrans for example
• we should have one way we're storing the data that we then use in all downstream results

AFTER:

env
    .packed_objects # a class that holds an array of PackedObject, and has helper functions for getting things like all the positions or all the radii of the currently packed objects

containers 
    .packed_objects # a subset of all packed objects, and only the ones that are in this container

ingredient
    .results # an array of [position, rotation]