bauer

Bayesian Estimation of Perceptual, Numerical and Risky Choice.

bauer is a PyMC-based Python library for fitting hierarchical Bayesian cognitive models to behavioural decision-making data. It covers magnitude comparison, psychometric functions, and risky choice — from simple Weber’s-law models to flexible noise curves and prospect-theory variants.

Key features

  • Ready-to-use model classes for magnitude comparison, psychometric functions, and risky choice — no need to hand-code PyMC models.

  • Hierarchical fitting by default: group mean + between-subject SD inferred jointly with subject-level parameters. Essential for typical trial counts (100–250 per condition).

  • Regression support via patsy formulas: e.g. regressors={'nu': 'C(condition)'} to let any parameter vary by experimental condition.

  • Posterior predictive checks with model.ppc(data, idata).

  • Full ArviZ integration: trace diagnostics, HDI plots, ELPD model comparison.

  • Included datasets: Garcia et al. (2022) magnitude/risk, de Hollander et al. (2024) dot-cloud and symbolic gambles.

Quick start

from bauer.models import MagnitudeComparisonModel
from bauer.utils.data import load_garcia2022

data = load_garcia2022(task='magnitude')
model = MagnitudeComparisonModel(paradigm=data, fit_seperate_evidence_sd=True)
model.build_estimation_model(data=data, hierarchical=True)
idata = model.sample(draws=1000, tune=1000)

Getting started

Tutorials

Reference

Indices and tables