Tutorial

This section uses a series of Jupyter Notebooks to explain how PyBLP can be used to solve example problems, compute post-estimation outputs, and simulate problems.

For a more concise, targeted, and opinionated tutorial, see the Demand Estimation Mixtape Session, a three day workshop taught by Jeff Gortmaker in 2024. In order, the days cover pure logit estimation, aggregate BLP estimation, and micro BLP estimation. Each day has slides and extensive coding exercises (along with Jupyter Notebook solutions) that use PyBLP.

In addition to the notebooks here and the above workshop, other examples can be found in the API Documentation.

• Logit and Nested Logit Tutorial
  • Theory of Plain Logit
  • Application of Plain Logit
  • Theory of Nested Logit
  • Application of Nested Logit
• Random Coefficients Logit Tutorial with the Fake Cereal Data
  • Theory of Random Coefficients Logit
  • Random Coefficients
  • Specification of Random Taste Parameters
  • Loading Data
  • Setting Up and Solving the Problem Without Demographics
  • Adding Demographics to the Problem
  • Restricting Parameters
• Supply Side Tutorial with Automobile Data
  • Loading Data
  • Setting up the Problem
  • Solving the Problem
• Micro Moments Tutorial with Automobile Data
  • Loading Data
  • Setting up the Problem
  • Setting up Micro Moments
  • Solving the Problem
  • Running the Main Counterfactual
• Post-Estimation Tutorial
  • Problem Results
  • Elasticities and Diversion Ratios
  • Marginal Costs and Markups
  • Mergers
  • Bootstrapping Results
  • Optimal Instruments
  • Optimal Micro Moments
• Problem Simulation Tutorial