Jomini

Historical battle simulation package for Python

Description

Jomini creates military simulations by using mathematical combat models. Designed to be helpful for game developers, students, history enthusiasts and -to some extent- scientists. You can mail me at [email protected] if you want to contribute.

Documentation will be available in the next update.

Programming language: Python

License: MIT License

Tags: Build Tools     Game Development     Library     Scientific     Mathematics     Python Modules     Python     Science    

Latest version: v0.1.4

README

Jomini v0.1.4

Jomini creates military simulations by using mathematical combat models. Designed to be helpful for game developers, students, history enthusiasts and -to some extent- scientists. You can mail me at [email protected] if you want to contribute.

To download: https://pypi.org/project/Jomini

Documentation will be available in the next update.

Lanchester Models 101

This package uses the combat models developed by Frederick William Lanchester, a.k.a one of the founding fathers of Operations Research.

• Lanchester Models are deterministic, which means the model will always yield the same result for the same input parameters.
• Lanchester Models view battles as an attrition model, therefore manuevers and sudden changes during the battle can not be represented.
• You might need to do some manual fine-tuning if you are not able to get quality parameters (rho, beta, engagement_width) from a data set.
• Despite the downsides, even the primitive models developed by Lanchester himself works wonders with the right parameters.

Code Sample: Lanchester's Linear Law

• The Linear Law is based on force concentration
• Good for modelling melee battles and unaimed fire (artillery, arquebus, handcannon etc.)

from jomini.Lanchester import Battle, LinearLaw

# Re-creating the Battle of Cerignola (AD 1503)
# In the actual battle, Spanish(red) lost 500 men while the French(blue) lost 4000 men
# Parameters rho, beta, engagement_width and time are manually fine_tuned
b = Battle(red=6_300, blue=9_000, rho=0.0800, beta=0.0100)
L = LinearLaw(b, engagement_width=100)
print(L.get_casualty_rates()) # Returns casualty rates 
print(L.get_casualties(time=7))
print(L.get_remaining(time=7))
print(L.simulate_battle(time=7))

Code Sample: Visiting models used in Jomini

• Square Law: Given equal power coefficients, the fighting power is proportional to the square of army size.
  • Good for modelling aimed fire (e.g Napoleonic line-battles)
• Logarithmic Law: Basically square law at a larger scale, used by Weiss to model the American Civil War
  • Good for modelling tank combat as well

from jomini.Lanchester import Battle, LinearLaw, SquareLaw, LogarithmicLaw, GeneralLaw

# Simulating a fictitious battle with each of the laws
b = Battle(red=20_000, blue=30_000, rho=0.0150, beta=0.0120)
Linear = LinearLaw(b, engagement_width=500)
Square = SquareLaw(b)
Log = LogarithmicLaw(b)
Generalized = GeneralLaw(b, engagement_width=500, p=0.450, q=0.600)

# If time is not specified, the battle goes on until one side is annihilated.
print(Linear.simulate_battle() + "\n")
print(Square.simulate_battle() + "\n")
print(Log.simulate_battle() + "\n")
print(Generalized.simulate_battle())