Multi-Stage Robust Decision Making (MSRDM)

Note that this package is not yet implemented.

Multi-Stage Robust Decision Making (MSRDM) is a decision-support framework under deep uncertainty.

Installation

MSRDM runs on Python 3.8 or higher.

Dependencies are numpy and pandas. Optional dependencies are matplotlib, seaborn, and plotly, which are required for plotting functions.

It can be installed by

pip install msrdm

Usage

HoU Analysis

from msrdm.uncertainty import RealUncertainParameter, EllipticUncertaintyRegion
from msrdm import hou_analysis

# Define uncertain parameters
w1 = RealUncertainParameter('w1', nominal=0, lower_deviation=1, upper_deviation=1, dog='greater')
w2 = RealUncertainParameter('w2', nominal=3, lower_deviation=4, upper_deviation=5, dog='greater')

# Define uncertain region
uncertainty_region = EllipticUncertaintyRegion([w1, w2])

# Define objective function
def func(x, w):
    assert x == 1 or x == 2
    return w[x]

# Conduct HoU analysis
result = hou_analysis(func=func, policies=[1, 2])

To-Do's

License

• BSD 3-Clause License