Pattern Discovery | Early Warning Signals | Regime Detection

Introduction | The General Causal Effect Identification Problem | The dosearch package | References

Using dynamite to generate panel data | Model with categorical responses | Model with time-varying effects

Default prior distributions in dynamite | Regression coefficients | Intercept | Standard deviation parameters | Correlation matrices | Parameters related to latent factors | Family-specific parameters | Defining priors of dynamite models

Introduction | The dynamic multivariate panel model | The dynamite package | Bayesian inference of seat belt usage and traffic fatalities | Causal effects in a multivariate model | Model construction | Defining response variables | Lagged responses and covariates | Time-varying and time-invariant effects | Group-level random effects | Latent factors | Multivariate responses | Number of trials and offset variables | Auxiliary response variables | Visualizing the model structure | Model fitting and posterior inference | User-defined priors | Extracting model fit information | Visualizing the posterior distributions | Missing data and multiple imputation | Prediction | Aggregated predictions and memory conservation | Summary | References | Details on latent factors

Using modified Stan code with dynamite | Using non-MCMC algorithms in dynamite