seedling · updated 2025-12-01 12:02:00

mcbayes

Hierarchical Bayesian analysis of R&D expenditure effects on GDP growth in Europe.

Approach

Statistical Rethinking (McElreath) workflow:

  1. DAG-based causal identification with do-calculus
  2. Hierarchical Bayesian regression
  3. NUTS sampler via PyMC
  4. ArviZ for diagnostics
  5. Interactive Plotly visualizations

Data

Eurostat API (2016-2022): - GDP by country - R&D expenditure

Why Bayesian

  • Quantified uncertainty (not just point estimates)
  • Hierarchical structure respects country-level variation
  • Prior knowledge incorporated explicitly
  • Full posterior for decision-making

Outputs

  • Marimo reactive notebook for exploration
  • Formal policy note with results
  • Interactive visualizations

Related

Applied example of Budget Constraints Drive Emergence - constraints and emergence in statistical modeling.