Multivariate Spatio-Temporal Models using Structural Equations
Make mesh for stream network
Simulate GMRF for stream network
Extract the (marginal) log-likelihood of a tinyVAST model
Make a RAM (Reticular Action Model)
Sample from predictive distribution of a variable
Construct projection matrix for stream network
Make a RAM (Reticular Action Model)
Multivariate Normal Random Deviates using Sparse Precision
Add mesh covariates to vertices and triangles
Rotate factors to match Principal-Components Analysis
Add predictions to data-list
Calculate conditional AIC
Classify variables path
Conditional simulation from a GMRF
Calculate deviance explained
Additional families
Get data
Get response
Integration for target variable
Make a RAM (Reticular Action Model) from a SEM (structural equation mo...
Parse path
Predict using vector autoregressive spatio-temporal model
print summary of tinyVAST model
Project tinyVAST to future times (EXPERIMENTAL)
Objects exported from other packages
Reload a previously fitted model
Calculate deviance or response residuals for tinyVAST
Simulate new data from a fitted model
Approximate spatial correlation
summarize tinyVAST
Extract covariance
Fit vector autoregressive spatio-temporal model
Control parameters for tinyVAST
Extract Variance-Covariance Matrix
Fits a wide variety of multivariate spatio-temporal models with simultaneous and lagged interactions among variables (including vector autoregressive spatio-temporal ('VAST') dynamics) for areal, continuous, or network spatial domains. It includes time-variable, space-variable, and space-time-variable interactions using dynamic structural equation models ('DSEM') as expressive interface, and the 'mgcv' package to specify splines via the formula interface. See Thorson et al. (2025) <doi:10.1111/geb.70035> for more details.