pyFIES

Python implementation of FAO's Food Insecurity Experience Scale (FIES) — the methodology behind UN SDG indicator 2.1.2 (prevalence of moderate-or-severe and severe food insecurity in the population).

pyfies is a from-scratch port of the R package RM.weights (Cafiero, Viviani, Nord), implementing the weighted Conditional Maximum Likelihood (CML) Rasch estimator, equating to FAO's 2014–2016 global standard, and probabilistic prevalence assignment along the latent food-insecurity trait.

Status: alpha. v0.1 covers the dichotomous Rasch model, equating, and prevalence estimation. The polytomous (partial credit) extension and full diagnostics suite are planned for v0.2.

Installation

pip install pyfies          # from PyPI (once published)
pip install -e ".[dev]"     # editable install with dev tooling

Requires Python 3.11+.

Quickstart

import pandas as pd
from pyfies import RaschModel, FAO_2014_2016, DEFAULT_FIES_ITEMS

# df has 8 columns named WORRIED, HEALTHY, FEWFOOD, SKIPPED,
# ATELESS, RUNOUT, HUNGRY, WHLDAY (1=affirmative, 0=negative, NA=missing)
X = df[DEFAULT_FIES_ITEMS].to_numpy()
w = df["sampling_weight"].to_numpy()

model = RaschModel().fit(X, sample_weight=w)
model.equate(reference=FAO_2014_2016)
result = model.prevalence()

print(f"Moderate or severe food insecurity: {result.moderate_or_severe:.1%}")
print(f"Severe food insecurity:             {result.severe:.1%}")

What it computes

1. Item severity parameters for each FIES question via weighted CML.
2. Person parameters per raw score (post-hoc MLE conditional on item parameters).
3. Equating to the FAO 2014–2016 global metric so prevalence rates are comparable across countries and survey rounds.
4. Prevalence rates at any latent-trait threshold via Gaussian-mixture probabilistic assignment.

Methodology references

• FAO. The Food Insecurity Experience Scale — Development of a Global Standard for Monitoring Hunger Worldwide. Technical Paper v1.1, 2016. link
• Cafiero, C., Viviani, S., Nord, M. (2018). Food security measurement in a global context: The Food Insecurity Experience Scale. Measurement, 116, 146–152. doi:10.1016/j.measurement.2017.10.065
• FAO. Voices of the Hungry / FIES landing page.

Numerical parity with RM.weights

pyFIES is validated against the reference R package on the data.FAO_country1..4 sample datasets shipped with RM.weights. To regenerate parity fixtures (R required, one-time):

Rscript scripts/generate_r_fixtures.R

Default tolerance: atol=1e-6 on item severities and prevalence rates.

Citing

If you use pyFIES in published research, please cite both this package and the underlying FAO methodology (Cafiero et al., 2018). A CITATION.cff file will be added at first stable release.

License

Apache 2.0.