Section dedicated to epidemiological modeling with climasus4r. Each card covers a family of analytical designs — DLNM, attributable burden, case-crossover, interrupted time series, multi-city pooling, spatial models, machine learning, and vulnerability — always connecting the fitting function to the corresponding diagnostics and visualizations.
| Model | Functions | Focus |
|---|---|---|
| 1 | sus_mod_dlnm(), sus_mod_plot_dlnm() | Exposure-response relationship with lag structure |
| 2 | sus_mod_af(), sus_mod_burden(), sus_mod_excess() | Attributable fraction, attributable burden, and excess events |
| 3 | sus_mod_casecrossover(), sus_mod_its() | Conditional case-crossover and interrupted time series |
| 4 | sus_mod_pool(), sus_mod_metaregression() | Multivariate pooling and meta-regression |
| 5 | sus_mod_ml(), sus_mod_plot_ml() | Machine learning and prediction |
| 6 | sus_mod_spatial_moran(), sus_mod_spatial_reg(), sus_mod_spatial_scan() | Autocorrelation, spatial regression, and cluster detection |
| 7 | sus_mod_spatial_bayes(), sus_mod_spacetime_bayes() | CAR/BYM models and spatio-temporal Bayesian models |
| 8 | sus_mod_vulnerability_index(), sus_mod_swot() | Vulnerability indices and SWOT analysis |
DLNM
Distributed Lag Non-Linear Models. Estimate exposure-response relationships with a lag structure between temperature and health outcomes.
View Tutorial →Attributable Burden & Excess
Calculate attributable fractions (AF), attributable number (AN), excess mortality, and disease burden imputable to climate exposure.
View Tutorial →Case-Crossover & ITS
Confounding controls in the time series: conditional case-crossover design and interrupted time series (ITS) analysis.
View Tutorial →Pooling & Meta-regression
Combine results from multiple cities with multivariate pooling, explore heterogeneity with meta-regression, and assess robustness with sensitivity analysis.
View Tutorial →Machine Learning
Apply XGBoost and other ML algorithms for health outcome prediction and identification of climate and socioeconomic predictors.
View Tutorial →Spatial Modeling
Spatial autocorrelation (Moran), spatial regression (frequentist SAR/CAR), and spatial scan statistics for detecting risk clusters.
View Tutorial →Spatial & Spatio-temporal Bayes
Bayesian BYM2/CAR models for spatial risk smoothing, spatio-temporal models, and exceedance probability via CARBayes/INLA.
View Tutorial →Vulnerability & SWOT
Build climate vulnerability indices (IPCC framework) integrating exposure, sensitivity, and adaptive capacity, and synthesize with SWOT analysis.
View Tutorial →