Add tools module and lattice_tools

simphony / simphony-common

The native implementation of the Simphony cuds objects

BSD 2-Clause "Simplified" License

7 stars 5 forks source link

Add tools module and lattice_tools #249

Closed kitchoi closed 8 years ago

kitchoi commented 8 years ago

Added lattice_tools as simphony.tools.lattice_tools

please comment if you think a different naming or hierarchy is better) Added tests for the lattice_tools
It takes a while because hypothesis tries to randomly construct many different primitive cells to catch edge cases; most of the time is spent on generating many primitive cell objects, not on calling lattice_tools Added hypothesis to dev_requirements.txt
- hugely useful for development (that's how I found out about #245)

Please play with it with your use cases.

Related to #240

kitchoi commented 8 years ago

Discussion with Adham led me to realise primitive cell configurations not included in the code. Please hold off review for now. Thanks!

kitchoi commented 8 years ago

Added features:

Parameter tolerance is added for controlling the numerical error tolerated
same_lattice_type is renamed to same_primitive_cell_config as the different primitive cell configurations can be applied for the same lattice type
the use of same_primitive_cell_config in more symmetric lattice types is replaced by more explicit calculations for enhanced computational performance
alternative primitive cell configurations not included in the PrimitiveCell module are now supported for body-centered cubic, body-centered tetragonal, body-centered orthorhombic and base-centered orthorhombic (please review and see if I still miss some)
guess_primitive_vectors is made more robust

Please review. Thanks.

kitchoi commented 8 years ago

A documentation (for sphinx) is not yet written. I believe it is better to write a section for a set of tools under the tools module (lattice, crystal) once they are finalised.

tuopuu commented 8 years ago

@kitchoi, I left a few minor comments between the lines. Nice looking code generally. This package will be very useful to crystallographers and lattice-Boltzmann enthusiasts. :+1:

kitchoi commented 8 years ago

Thanks @tuopuu! All fixed. Please let me know if this is good to merge. Thanks.

tuopuu commented 8 years ago

@kitchoi, looks good. :+1:

I would have liked to still improve the docstrings in same_primitive_cell_config(), but maybe it's not worth the time. :smile: It's not so easy to define what similarity there means without going too far into mathematics. Anyway, it should be intuitive enough what a comparison between cosines and length ratios here does.