encode vs. decode symprec, angle_tolerance, and primitive options

[sgbaird]

No description provided.

[sgbaird]

Related:

• Get default feature ranges based on conventional cells #96
• Consider using Wyckoff positions instead of atomic sites #94
• consider implementing fit method #89
• Generated, decoded structures all seem to be of P(1) symmetry #79
• Does increasing symprec in pymatgen decrease number of sites? #65
• Maybe use get_sites_in_sphere to calculate pairwise distance matrix, so always 52 sites #11

[sgbaird]

#65

[sgbaird]

Need to add tests for this functionality

[sgbaird]

Something strange with the final coveralls test (showed up as "decreased to zero"). lemurheavy/coveralls-public#586

Keep an eye out to see if this behavior persists.