Skip to contents

Compute impulse response functions for a VAR(P)

Source: R/computeirf.R
computeirf.Rd

Recursively computes IRFs given an impact matrix and VAR coefficient array. Optionally cumulates responses for selected variables.

Usage

computeirf(Psi, Phi, H, cum)

Arguments

Psi

Impact matrix, dimension N x R.

Phi

Array of VAR coefficient matrices, dimension N x N x P.

H

Horizon (number of periods) for which to compute IRFs.

cum

Logical scalar or logical vector of length N. If TRUE for variable i, the IRF is cumulated via cumsum. A single value is applied to all variables.

Value

An array:

irf

Array of impulse responses, dimension H x N x R. Row names are labelled 0 to H-1.

Examples

Psi <- matrix(c(1, 0.5), nrow = 2)
Phi <- array(0, dim = c(2, 2, 1))
Phi[, , 1] <- matrix(c(0.4, 0.1, 0, 0.3), 2, 2)
computeirf(Psi = Psi, Phi = Phi, H = 4, cum = FALSE)
#> , , 1
#> 
#>    [,1]   [,2]
#> 0 1.000 0.5000
#> 1 0.400 0.2500
#> 2 0.160 0.1150
#> 3 0.064 0.0505
#>