Setup | Helper functions | Wide-format conversion for Apollo | Apollo probability function factory | Apollo starting-values vector | Fit both packages, return comparison table | 1 · Parameter recovery | 2 · Timing benchmark | 2a · MLE estimation timing | 3 · Demand simulation: 1% price increase | Policy specification | 3a · Apollo apollo_prediction() approach (convergence check, n=500, J=5) | 3b · Demand simulation timing (rmdcev vs Apollo, n=500) | 4 · Results summary | MLE estimation | Demand simulation (100 draws per individual, 3 replicates, n=500) | Demand change convergence (n=500, J=5, 1% price increase, same additive policy) | Session info