search

mountaindust / Planktos

ABM framework for dispersal modeling
https://planktos.readthedocs.io
GNU General Public License v3.0
5 stars 3 forks source link

Need to work on data output #37

Closed mountaindust closed 3 years ago

mountaindust commented 3 years ago

This has been neglected somewhat... should output data from simulations in a format that is friendly to opening in VisIt or some kind of data analysis software such as R (e.g., csv).

mountaindust commented 3 years ago

The way forward on this is probably to create a save_all routine in the environment class which saves everything in a human-readable format that can also be read back into Planktos at a later date. We should look into using .vertex files that can be read into VisIt the same way meshes are to plot the agent positions over time. Then there can be a .out file of some sort for the environment and each associated swarm which lists attribute values. Swarms should be given a "name" attribute for easier recall of which swarm was for what, and the flow field should not be saved. Instead, the location of the fluid data or type of analytic flow field should be recorded so that it can be auto-loaded again later without taking up additional space.

mountaindust commented 3 years ago

It's going to be way too much to save everything in a way that can completely reload a session. There are just too many ways of getting fluid velocity fields into Planktos (a good thing!), parameters that could be involved in generating those flows, etc. At some point, we're just going to have to assume that the user is saving the py or ipynb file of what they were doing and could reconstruct the environment. So then, the main thing to concentrate on are the data that are generated specifically by Planktos during a simulation.