search

mountaindust / Planktos

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

RK45 step solver option #13

Closed mountaindust closed 3 years ago

mountaindust commented 6 years ago

Currently, all agent movement is Euler. It would be nice to implement a runge-kutta 45 solver to get more accurate results.

mountaindust commented 3 years ago

This only makes sense when there is no noise, which will not be the case most of the time. It's main use-case is if we are going to implement FTLE as part of this package. More pressing is to get better interpolation methods as an option, such as tricubic or quadcubic. There's a good looking tricubic algorithm at eqtools, and quadcubic is available through ARBInterp, but it's not yet peer reviewed. Closing this in favor of two new issues, one to address interpolation accuracy and another to more specifically address solving the Ito and Stratonovich equations governing simple IBMs with no agent-to-agent interaction.