Currently, AGNI is called as a subprocess through its command line interface. This means that each call to the model needs to re-allocate memory and load data from the disk.
It would be more efficient/faster to use JuliaCall to access the model, where the data can be held in memory by PROTEUS for the entire duration of the simulation.
Currently, AGNI is called as a subprocess through its command line interface. This means that each call to the model needs to re-allocate memory and load data from the disk.
It would be more efficient/faster to use JuliaCall to access the model, where the data can be held in memory by PROTEUS for the entire duration of the simulation.
https://juliapy.github.io/PythonCall.jl/stable/juliacall/