Why model invasions as relational events? | Ingredients | A simulated invasion record | Why ordinary case-control sampling cannot see ENSO | Time-shifted sampling | Recovering the drivers | Where to go from here | References

1. Load a bundled REM dataset | 2. Build candidate specifications | 3. Compare by AIC | Multiple controls per case | 4. Inspect coefficients of a chosen specification | 5. Cross-implementation guarantee | References

Overview | Causal Poisson regression | Exhaustive search with chi-square test | Stepwise search | Causal logistic regression | Causal GAMs for nonlinear relationships | Larger example: 5 covariates | Summary of recommendations | Reference

Overview | Causal Poisson regression | Exhaustive search with chi-square test | Stepwise search | Causal logistic regression | Causal GAMs for nonlinear relationships | Larger example: 5 covariates | Summary of recommendations | Reference

Endogenous reciprocity | Time-varying global covariates | Composing endogenous and global | Caveat

US state distance matrix | Defining a non-linear effect | Simulating events with exogenous covariates | Recovering the effect with a GAM | Plotting estimated vs true effect

Simulating actor covariates | Simulating event sequences

What this package does | Loading a tree (offline demo) | Loading a real atlas tree (requires network) | Browsing available trees | Pointing at a fork or local mirror | Provenance and reproducibility | Where to learn more

Simulating Data | Estimation | Results