simulate_conditional_forecasts: Simulate paths from conditional forecast distributions

In APRScenario: Structural Scenario Analysis for Bayesian Structural Vector Autoregression Models

Simulate paths from conditional forecast distributions

Description

Simulate paths from conditional forecast distributions

Usage

simulate_conditional_forecasts(mu_y, Sigma_y, varnames, n_sim = 1000)

Arguments

mu_y	Array (n_state × 1 × n_draws): conditional forecast mean
Sigma_y	Array (n_state × n_state × n_draws): conditional forecast variance
varnames	Character vector of variable names (length = number of variables)
n_sim	Number of simulations per draw

Value

Array of dimensions (n_state x n_sim x n_draws) of simulated draws with named rows

Note

Users should set their own seed before calling this function if reproducible results are desired.

Examples

# Example with simulated data
# Create example data dimensions
n_var <- 3
h <- 2
n_draws <- 5
n_state <- n_var * h

# Simulate conditional forecast means and covariances
set.seed(123)
mu_y <- array(rnorm(n_state * 1 * n_draws), dim = c(n_state, 1, n_draws))
Sigma_y <- array(0, dim = c(n_state, n_state, n_draws))
for (d in 1:n_draws) {
  temp_cov <- matrix(rnorm(n_state^2), n_state, n_state)
  Sigma_y[,,d] <- temp_cov %*% t(temp_cov) + diag(n_state) * 0.1
}

# Variable names
varnames <- c("GDP", "CPI", "FFR")

# Simulate conditional forecasts
sims <- simulate_conditional_forecasts(mu_y, Sigma_y, varnames, n_sim = 50)
print(dim(sims))
print(rownames(sims)[1:6])

APRScenario documentation built on Dec. 22, 2025, 1:06 a.m.