Discussion on the model_data repo

A few questions on the model_data repo that I hope one of you can answer:

README.md:

• Should “make pdf“ be “make latexpdf“?

interaction.py:

• Could we rename “intr“ to “interaction”?

body.py:

• Should “SomeMaterialClass” be “some_material_class”?
• So there are mandatory (e.g. position) and optional variables (e.g. temperature). Should I add a list of optional variables that are used in MercuryDPM (e.g. liquid_content) , or should that be done later?
• Why is “color“ a property of “shape“? Maybe put it directly in “body”?
• Should “material“ be stored as an id, similar to the body “id”’s in “interaction”?
• I don’t like the “lifespan“ variable – it changes as time evolves. How about storing creation_time instead?
• Classes like “normal_linear“: are these used to store history-dependent parameters of the interactions? Should they have a common base class (e.g. SomeMaterialParams) and should the interaction class contain one of these objects (i.e. add “material_params: type(SomeMaterialParams) = SomeMaterialParams()” to “intr”)?

model.py:

• There should be a correlation between the models and the material properties. Maybe the model should specify which child of “SomeMaterialClass“ and “SomeMaterialParams” is used.

scene.py:

• Should the scene class contain a model and a list of materials, bodies, interactions? Or is there be an overarching dem class that contains everything?
• Should bodies be split into moving particles from fixed walls? Alternatively, should there be a property that distinguishes moving and fixed bodies?
• Add timeMax? Add periodic boundaries? Add inflow/deletion regions? Add name (of output files)?