Introduction | Generalized Linear Mixed Models | Adaptive Gauss-Hermite quadrature for GLMMs | Model fitting | Examples | Troubleshooting | Appendix: derivation of PIRLS

Introduction | Formula module | Objective function module | Nonlinear optimization module | Output module | Conclusion | Modularization examples

Introduction | Background | Construction of the Relative Co-factor | Extracting model components | References

Autoscale Mechanism | References

Overview | Standard Matrix Multiplication (%*%) | Kronecker Product (%x%) | Elementwise Multiplication with Recycling (*) | Example of Age-Structured Transmission with Elementwise Matrix Multiplication | Multiplying a Sparse Matrix with a Dense Matrix | Extracting Sparse Representation | Small Example of Sparse Matrix Multiplication | Example of Sparse Matrix Multiplication to Model Time-Varying Transmission Rate | Summary of Matrix Operations | When to Use Each

Potentially Useful | Logging and Verbosity | Output Data Formatting (for Scalar-Only Models) | Numerical Tolerances | Advanced | Simulation and Reproducibility | Time-Restricted Functions | Engine Control | Deprecated | No Longer or Not Yet Used

Hello World | Creating a Simulator | Generating Simulations | Processing Results

Decomposition of State-Dependent Rates | Example -- SIR | Product Models | Example -- SIR times age | Example -- SIR with multiple I-boxes times age | Example -- SIR with multiple I-boxes and partial immunity times age | Example | Bookkeeping (in progress)

Introduction | Utilities and Packages | Initial State Vectors | Model Parameters | Derived State Vectors | Per-Capita Transmission Matrix | Per-Capita Force of Infection | Absolute Flows | Allocation Matrices | Euler Step Update | Model Specification | Simulation and Plots | Limitations

Hello, World | Step 0: set up simulator and generate 'data' | Step 1: add calibration information | Step 2: do the fit | Step 3: check the fit | Statistical Model

Introduction | Setup and Example Models and Data | Naming Conventions | What does macpan2 mean? | Why do most of the functions start with mp_? | Why do some of the functions start with mp_tmb_? | Calibration Details and Outputs | How does macpan2 fit models to data? | How does macpan2 derive confidence intervals? | What likelihood and prior distribution assumptions does macpan2 make? | Can I get a new model specification with fitted parameters as defaults, from a calibrated calibrator object? | Plotting Simulated Trajectories | How can I plot simulated trajectories? | How can I plot a calibrated model with the observed data used in the calibration? | Calibration Troubleshooting | Should I worry about NA/NaN function evaluation warnings? | What should I do if I do not converge? | How do I increase the number of iterations that the optimizer takes? | Parameter Estimation and Uncertainty | What can I do when my confidence intervals include zero for a parameter that should be positive? | Can I weight the importance of trajectories when fitting to more than one of them? | Dynamical Instabilities | What is the best way for me to keep my state variables from going negative? | Initial Conditions | My data start mid-epidemic: how do I know what the initial number of infectious individuals is? | Coding Style | What does |> mean? | Why do you put weird symbols at the start of lines? | Time Variation of Parameters | How can I use a known time series of values? | What if my time-series has missing values? | What can I do if I do not know the functional form of time-variation? | Under Reporting, Delayed Reporting, and Total Reporting | How should I account for under-reporting in macpan2? | How should I account for delayed reporting as well as under-reporting in macpan2? | Can macpan2 use hidden Markov models to account for reporting bias? | Stochastic Simulation | How can I set the seed for the random number generator? | Why do I keep getting the same result even though my model has randomness? | What are the mathematical details of the Euler multinomial process error model? | Does macpan2 implement the Gillespie algorithm? | More Process Error (incomplete)

Setup | One Scarlet Fever Outbreak in Ontario | SIR Model | Modify Model for Reality | Preparing the Data | Set up the Optimizer | Run the Optimization | Examine the fit | Fixing the Optimization Problem | Learning the Functional form of Time Variation in Transmission (New!) | References

Calibration Defaults | Customizing the Objective Function | Use Cases | Priors on model parameters | Likelihoods on model parameters | Fixed distributional parameters | Fitting Distributional Parameters | Distributional Parameter Transformations | Priors on distributional parameters

Information Processing | Modularity | Engine-Agnostic Model Specification Language | Engine-Specific Model Specification Languages | General Dynamic Simulation with TMB | Model Library | Calibration | Specifying Data to Fit | Specifying Distributional Assumptions | Specifying Parameters to Fit | Alternative Trajectory Solvers | Euler Step | Mean Euler-Multinomial (a.k.a. Hazard Correction) | Runge-Kutta 4 | Time-Varying Parameters | Vectors in the TMB Engine | Model Structure and Bookkeeping | Structure in Expressions | Constructive Descriptions of Model Structure | Structured Vectors | Structure Encourages Reparameterization | Alternative Engines | Combining Expression Lists

Amuse bouche: a structured SIR model | Appetizer: specifying the basic SIR model | Main course: expanding the basic SIR with additional structure | Dessert: understanding model simulation in macpan2

'Hello, World': an easy (??) calibration exercise | Step 0: set up simulator and generate 'data' | Step 1: add calibration information | Step 2: do the fit | A slightly harder, more realistic example | Computing and plotting confidence and prediction intervals | Hamiltonian MC | Alternative confidence intervals | Irregular data; multiple data streams | 'Hello, World' the hard way | Step 0: recreate the simulator | Step 1: add observed data and slots for history etc. | Step 2: collect simulated values | Step 3: set up and compute objective function | Step 4: declare and/or transform parameters to be optimized, set starting values | Step 5: do the fit | Measles Data -- In-Progress | Challenging Logistic Variation in Transmission Rate | References

Prerequisites | Matrices | Parameters | Trajectory Simulation | Expressions | Parse Tables | Assignment | Assignment Parse Tables | Literals | Function Definitions | Objective Function

The Problem | Definition of an Elementwise Binary Operator in the C++ Engine | Forcing a Binary Operator in R to have these Properties | Step 1 | Step 2 | Step 3 | Implementation and Examples

Introduction | Basics of the macpan2 Object Oriented Framework | Constructing and Using Objects | Defining Classes | Details | Objects | Class Definitions | Inheritance | Principles | Small Classes | Avoid Modifying Well-Tested Classes | Linear Inheritance | Shallow Inheritance Hierarchy | Balance Regeneration with Consistency | Patterns | Alternative Classes | Argument Fields | Static Fields | Standard Methods | Composition | Refresh Methods | Private Methods | Object Editing | Method Caching

Finding Example Models | Using Examples | Modifying Examples

Introduction | Mixed-model-specific issues | Terminology | Time-consuming computations | Related packages | huxtable + broom.mixed | dotwhisker + broom.mixed | Capabilities

Example: Orobanche/overdispersed binomial | Example: reed frog size predation | Newer stuff | Technical details

Branching Flows and Process Error | Internal Design | ChangeComponent() classes | The change_frame() method | The flow_frame() method | ChangeModel() classes | UpdateMethod() classes | Relationship to Ordinary Differential Equation Solvers | Euler | Runge Kutta 4 | Euler-Multinomial | Hazard | Linearizing at Each Time-Step | Hazard in models including more than unbalanced per-capita flows

Baseline SIR Model | Piecewise Time Variation | Calibrating Time Variation Parameters | Radial Basis Functions for Flexible Time Variation (In-Progress) | Calibration | References

Baseline SIR Model | Piecewise Time Variation | Calibrating Time Variation Parameters | Radial Basis Functions for Flexible Time Variation (In-Progress) | Calibration | References

Choosing a Debugging Technique

Introduction | Notation and Theory | Online Forecasting | Eigenvector State Initialization | Priors on R0 | Regenerating Versus Reparameterizing Models

Stable | Mature Draft | Working Draft | Stub

overview | setting calc.derivs = FALSE | choice of optimizer | changing optimizer tolerances | parallelization/BLAS | other packages

Introduction | Installation | Quick start (for the impatient) | Basics | Incorporating random effects

Motivation | Using Optional Arguments in margins() | Average Marginal Effects and Average Partial Effects | Using the at Argument | Subsetting | Marginal Effects Plots | The plot() method for "margins" objects | More advanced plotting | Interpreting Interactions with Marginal Effects | Interactions in OLS | ggplot2 examples | Interactions in Logit | References

OLS marginal effects | Stata | R | OLS marginal effects with interaction | OLS marginal effects with factor term | OLS marginal effects with squared term | OLS marginal effects with squared term (but no first-order term) | Logit effects on log-odds and probability scales | Logit effects with factor variable on log-odds and probability scales | Logit with interaction on probability and Log-Odds scales | Probit effects on latent and probability scales | Poisson effects on latent and probability scales