Forecasting#
Task-oriented examples for forecasting workflows, including Fisher matrices, DALI expansions, systematic parameter biases, and Laplace approximations.
For end-to-end workflows and larger examples, see the derivkit-demos repository.
Fisher
Fisher matrix forecasts, parameter uncertainties, and posterior contours.
Fisher bias
Parameter shifts from systematic model mismatch.
DALI
Higher-order likelihood expansions and non-Gaussian posterior contours.
DALI bias
Non-Gaussian MAP and posterior mean shifts from systematic mismatch.
Laplace approximation
Gaussian approximation around a posterior peak and its contours.
Gaussian Fisher
Fisher matrix including parameter-dependent covariance.
X–Y Gaussian Fisher
Fisher forecasts with uncertainty in both inputs and outputs.