pyfair

Factor Analysis of Information Risk (FAIR) model written in Python.

This package endeavors to create a simple API for automating the creation of FAIR Monte Carlo risk simulations.

This is based on the terms found in:

"Open FAIR" is a trademark of the Open Group.

Installation

pyfair is available on PyPI. To use pyfair with your Python installation, you can run:

pip install pyfair

Documentation

Documentation can be found at the Read the Docs site.

Code

import pyfair

# Create using LEF (PERT), PL, (PERT), and SL (constant)
model1 = pyfair.FairModel(name="Regular Model 1", n_simulations=10_000)
model1.input_data('Loss Event Frequency', low=20, mode=100, high=900)
model1.input_data('Primary Loss', low=3_000_000, mode=3_500_000, high=5_000_000)
model1.input_data('Secondary Loss', constant=3_500_000)
model1.calculate_all()

# Create another model using LEF (Normal) and LM (PERT)
model2 = pyfair.FairModel(name="Regular Model 2", n_simulations=10_000)
model2.input_data('Loss Event Frequency', mean=.3, stdev=.1)
model2.input_data('Loss Magnitude', low=2_000_000_000, mode=3_000_000_000, high=5_000_000_000)
model2.calculate_all()

# Create metamodel by combining 1 and 2
mm = pyfair.FairMetaModel(name='My Meta Model!', models=[model1, model2])
mm.calculate_all()

# Create report comparing 2 vs metamodel.
fsr = pyfair.FairSimpleReport([model1, mm])
fsr.to_html('output.html')

Report Output

Serialized Model

{
    "Loss Magnitude": {
        "mean": 100000,
        "stdev": 20000
    },
    "Loss Event Frequency": {
        "low": 20,
        "mode": 90,
        "high": 95,
        "gamma": 4
    },
    "name": "Sample Model",
    "n_simulations": 10000,
    "random_seed": 42,
    "model_uuid": "2e55fba4-c897-11ea-881b-f26e0bbd6dbc",
    "type": "FairModel",
    "creation_date": "2020-07-17 20:37:03.122525"
}