Tips and common issues | Non-standard distributions | Operator overloading | Initialising with NA | Wrapping arrays in AD() | No if statements on parameters | Byte compiler side effects | Further reading

Smooth transition probabilities | Smooth density estimation | Markov-switching GAMLSS | References

Calling other distributions | A complete example | Tests | Numerical stability | Branching without breaking the tape | Zero-modified distributions | Reparameterised variants | Distributions to read | Before you are done | When the cdf has no closed form | When RTMB has no AD version

Continuous distributions | Discrete distributions | Multivariate distributions | Copulas

Example 1: Random regression with non-Gaussian data | Example 2: Non-standard random GLM for count data | Example 3: Distributional regression with penalised splines | Example 4: Zero inflation | Example 5: Copulas | Example 6: Animal movement with a circular-linear copula | Example 7: Multivariate stochastic volatility

The generator matrix | From the generator to transition probabilities | Likelihood | Example 1: two states | Simulating data | Writing the negative log-likelihood function | Fitting the model | Results | Example 2: three states | References

Likelihood | MMPP | MMMPP | Example 1: Markov-modulated Poisson processes | Setting parameters | Simulating an MMPP | Writing the negative log-likelihood function | Fitting the model | Results | Example 2: Markov-modulated marked Poisson processes | Simulating an MMMPP | References

Example: stochastic volatility model | Simulating data | Writing the negative log-likelihood function | Fitting an SSM to the data | Results | Comparison to Laplace approximation | References

Introductory example: Homogeneous HMM | Generating data from a 2-state HMM | Inference by direct numerical maximum likelihood estimation | Visualising results | References

Multiple data streams | Covariate effects | Covariate effects in the state process | Simulation example | Fitting an HMM to the data | Covariate effects in the state-dependent process | Simulation | Writing the negative log-likelihood function | Fitting a Markov-switching regression model | Visualising results | A special covariate: Seasonality | References

Complete pooling | Generating data | Writing the negative log-likelihood function | Estimating the model | Partial pooling | Random effects | Writing the joint negative log-likelihood

Simulation example | Setting parameters | Simulating data | Writing the negative log-likelihood function | Fitting an HSMM (as an approximating HMM) to the data | Results | Real-data application | Fitting an HSMM (as an approximating HMM) to the muskox data | References