We are cleaning up the code for this reporsitory. To reproduce the result from the paper, please use this branch:

https://github.com/ndgnuh/RiceBPH/tree/paper-replicate

This is a model to simulate the rice-brown plant hopper dynamic. The goal of this model is to research the effect of cultivated flower on the spread inside rice fields of brown plant hopper (a dangerous rice pest).

This model is implemented using the Julia language, please install it from this link. We recommend using version Julia 1.9.

After that, setup the project's dependencies:

julia --project -e 'using Pkg; Pkg.resolve(); Pkg.instantiate()'

[Optional] Build a "system image". A system image contains compiled code, so that Julia does not have to recompile everything everytime it runs. TLDR: this reduces latency.

julia --project scripts/build_system_image.jl

As far as I know, you need a working C compiler for this step. If not, try installing Julia 1.9. The 1.9 version has binary caching so it should have less latency than previous versions.

Here we show some simple examples. The running scripts are placed in the scripts/ folder. If you wish to customize, all of the scripts have --help flags, which lists all the options.

Model exploration:

julia --project scripts/main.jl configs/exploration.toml

# If you built a system image when setting up
julia -J ricebph.sys.so --project scripts/main.jl configs/exploration.toml

Render a video:

julia --project scripts/main.jl configs/video.toml
julia -J ricebph.sys.so --project scripts/main.jl configs/video.toml

We setup tests to guarantee that our model is fully reproducible (see test/runtests.jl).

First replicate the data:

julia --project scripts/replicate_data.jl

Output data will be written to outputs.

And then replicate the figures:

julia --project scripts/replicate_figures.jl

The figures will be stored in figures.

If the replication is too slow, try using multi-processing. It's really simple:

# Eight processes
julia -p 8 --project scripts/replicate_data.jl