Sketch of code architecture

[EqualAPriori]

Key decisions:

• forcefield object handles writing to common language
• forcefield uses (nm), (kT), but also stores temperature of coarse graining
• forcefield is layerable, but add in support for PFTS defaults, as well as checks to ensure PFTS-simulatability
• exporters/importers handle reading/writing from sim and PolyFTS formats.
• topology: include a new class/data structure for specifying molecule architecture, that is in addition to
• system: very minimal core, will largely need a sim system object and a polyFTS system object
• ** in the future: maybe the "export/import" to sim should be built into the sim package??? **

--- details
exporting to MD,sim,FTS:

• export_wrapper(forcefield) (k*coarse-grainingT)
• import_wrapper( sim-->forcefield )
• takes a MD system --> create PolyFTS topology

exporting to our own format:

• forcefield.export ?

forcefield

• data: all the parameters
• functions:
• energy (kT), units (nm), but also store T of coarse-graining
• layerable (default to supporting full sim capabilities), but then add in checks on how to handle repeated potentials -->
• --> how to get bead types , check that smearing lengths are consistent
• --> how to get the bond type (for PolyFTS)
• --> defaults?

topology

• data:
  • bead names, bonds (conventional MD)
  • chain statistics determiner/guesser, for a first recommendation
  • have general tree/architecture class (FTS)
• functions:
  • mapper

system:

• data: core variables (T, P, maybe box, number of molecules)

• then "misc" classes

• separate MD and FTS system objects for storing MD and PolyFTS parameters. kT? units?

• functions

• kT? units?

Draft system file:

#system.yaml

core:
- T:
- P:
- contents:
    - DS 3
    - DS1 7
    - ...


MD:

FTS:

Three ways to specify a molecule:

- name: molname1
  def: filename
- name: molname2
  beads: [list]
  bonds: [list of pairs]
- name: molname3
  architecture: 
    all_the_comb_PFTS_specifications: data