A package for broadCASTing epidemiological and ecological models over META-populations.

MetaCast is a python package for broadcasting epidemiological and ecological ODE based models over metapopulations (structured populations). Users first define a function describing the subpopulation model. MetaCast's users then define the dimensions of metapopulations that this subpopulation is broadcast over. These dimensions can be flexibly defined allowing for multiple dimensions and migration (flows) of populations between subpopulations. In addition to the metapopulation suite MetaCast has several features. A multinomial seeder allows users to randomly select infected stages to place an infected population in based on the occupancy time of infected states. MetaCast's event queue suite can handle discrete events within simulations, such as movement of populations between compartments and changes in parameter values. Sensitivity analysis can be done in MetaCast using parallelisable Latin Hypercube Sampling and Partial Rank Correlation Coefficient functions. All of this makes MetaCast an ideal package not only for modelling metapopulations but for wider scenario analysis.

Python 3.10 and pip. Package requirements:

• numpy >= 1.26.3
• pandas >= 2.1.4
• scipy >= 1.11.4
• pingouin >= 0.5.4
• tqdm >= 4.66.1
• dask >= 2024.2.1
• distributed >= 2024.2.1

For running demonstration jupyter notebooks

• bokeh >= 3.3.4
• seaborn >= 0.13.2
• jupyter >= 1.0.0

Note this should also install required packages.

pip install metacast

See jupyter notebooks in demonstration directory of the projects GitGub repository: https://github.com/m-d-grunnill/MetaCast/tree/main/demonstrations

https://metacast.readthedocs.io/en/latest/index.html

Currently, MetaCast's lhs_and_prcc function only support latin hypercube sampling with uniform distributions.

If you wish to contribute to this project please see the CONTRIBUTING.md file.