Penalized Estimation of Multiple-Subject Vector Autoregressive Models

Simulate, estimate, and forecast vector autoregressive (VAR) models for multiple-subject data using structured penalization. Decomposes dynamics into shared (common) and subject-specific (unique) components via adaptive LASSO with FISTA optimization. Supports cross-validation and extended BIC model selection and subgroup detection, and time-varying parameters.

multivar

Penalized estimation of multiple-subject vector autoregressive (VAR) models. Estimates shared (common) and subject-specific (unique) network dynamics across multiple individuals using structured penalties with FISTA optimization.

Installation

# From GitHub
devtools::install_github("zackfisher/multivar")

Quick start

library(multivar)

# Simulate data: 2 subjects, 5 variables, 50 timepoints
sim <- multivar_sim(k = 2, d = 5, n = 50,
                    prop_fill_com = 0.1, prop_fill_ind = 0.1,
                    lb = 0.1, ub = 0.5, sigma = diag(5))

# Fit model
model <- constructModel(data = sim$data)
fit <- cv.multivar(model)

# View estimated dynamics
print_dynamics(fit)

Features

• Adaptive LASSO with debiased initial estimation
• Cross-validation and eBIC model selection
• Subgroup detection
• Parallel cross-validation

References

Fisher, Z. F., Kim, Y., Fredrickson, B. L., & Pipiras, V. (2022). Penalized estimation and forecasting of multiple subject intensive longitudinal data. Psychometrika, 87(4), 1377–1404.