Overview | 1. Forecast accuracy benchmark | Results | Comparison with published benchmarks | 2. Calibration diagnostics | Key finding | Recalibration | Conformal prediction | 3. Sequential detection | Limitation | 4. Fitness decomposition | 5. Multi-pathogen applicability | References

The resource allocation problem | Expected Value of Information | Adaptive allocation via Thompson sampling | Detection horizon | Sequential detection with controlled false alarms | Surveillance dashboard | Comparison with existing approaches | References

The confounding of transmissibility and immune escape | Constructing the immunity landscape | Fitness decomposition | DMS-informed early detection | Selective pressure as an early warning signal | Practical considerations | References

Why calibration matters | PIT diagnostics on real CDC data | Recalibration | Conformal prediction intervals | Adaptive conformal inference | Proper scoring rules | References

Motivation | Example: synthesizing patient records | Parametric synthesis (Gaussian copula) | Validation | Comparing methods | Privacy risk assessment | Downstream model fidelity | Privacy-utility trade-off

Overview | Example: COMPAS recidivism prediction | Create fairness data | Compute fairness metrics | Generate fairness report | Visualize disparities | Group-wise calibration | Threshold optimization | Intersectional analysis

Overview | Load data | Collapse rare lineages | Fit MLR model | Growth advantages | Frequency trajectories | Forecast | Emergence detection | Sequencing power | Session info

Motivation | Data source | Setting up the design | Sequencing inequality | The bias problem: weighted vs naive | Optimal resource allocation | Delay correction and nowcasting | Combined correction | Key takeaways | Reproducibility

The right-truncation problem | Estimating the delay distribution | Reporting probability | Nowcasting | Combined design + delay correction | Choosing a delay distribution

Why allocation matters | The three objectives | Example | Optimize for minimum MSE | Compare all strategies | With minimum coverage constraints | Choosing an objective

The problem | What survinger does | Quick start | Comparing weighted vs naive estimates | Optimizing resource allocation | Delay correction | Combined correction | Detection power | How survinger differs from phylosamp | Next steps

Overview | Setup | Engine 1: MLR | Engine 2: Piantham | Comparing fit statistics | Backtesting | Scoring | Model ranking | Visualization | When to use which engine | Hierarchical MLR

Overview | Preparing data | Fitting a model | Extracting growth advantages | Visualizing the fit | Forecasting | Detecting emerging lineages | Next steps

Overview | Step 1: Load and prepare data | Step 2: Collapse rare lineages | Step 3: Fit model | Step 4: Growth advantages | Step 5: Identify emerging lineages | Step 6: Forecast | Step 7: Assess sequencing needs | Step 8: Extract tidy results | Summary