Paper figures#

The 2020 JCIM figures are drawn from figshare project 73545. This page is the same calls on the dumps that are in the tree. tests/dseams/test_doc_snippets.py runs the Python block on this page. Python is pip install 'pydseamslib==2.14.0'. The signature cage count and vertex total come from dseams.userframe.

figure

dump in this tree

call

result

cubic ice, CHILL+

mW_cubic.lammpstrj frame 1, type 1

chill_plus()

4096 cubic

cubic ice, cages

the same frame

cages()

n_ic 4096

sI 512cages

genice_sI.lammpstrj, type 1

cages_by_signature("512")

2 cages, 40 atoms

mixed frame

exampleTraj.lammpstrj, type 2

cutoff CHILL+ versus cages

9 interfacial clathrate, no complete cage

The R and Julia fronts also report 4096 cubic on the first row. They call libseams_lang with the cutoff neighbour list, the same default as Frame.chill_plus. On this lattice that list and the mutual four-neighbour list agree.

The nucleation, nanotube, monolayer, and in-plane RDF figures need the figshare dumps. Point the same functions at those files.

import pydseams as ds

ice = ds.read("input/traj/mW_cubic.lammpstrj", frame=1, atom_type=1)
print(ice.chill_plus().cubic, ice.cages().n_ic)
assert ice.chill_plus().cubic == 4096 and ice.cages().n_ic == 4096

from dseams import userframe

guest = userframe.read("input/traj/genice_sI.lammpstrj", atom_type=1)
print(len(guest.cages_by_signature("512")))
assert len(guest.cages_by_signature("512")) == 2
assert guest.cages_by_signature("512").n_atoms == 40