Same quantity as irfun: the orthogonalised impulse response of a
VAR(p), the MA recursion Phi_0 = I, Phi_h = sum_{j=1..min(h,p)} A_j
Phi_{h-j}, with shocks orthogonalised by the lower Cholesky factor P of
Sigma_u, so Theta_h = Phi_h P. Despite the name this is not "variable
importance"; it is the VAR impulse response.
Usage
Varimp(coef, sigma_u, horizon = 20, shock_var = 0)
Arguments
- coef
m by (1 + m p) coefficient matrix from Varest.
- sigma_u
m by m residual covariance.
- horizon
Periods ahead.
- shock_var
0-based index of the shocked variable.
References
Lutkepohl, H. (2005). New Introduction to Multiple Time
Series Analysis. Springer, Ch. 2.3. doi:10.1007/978-3-540-27752-1.
Examples
Varimp(matrix(c(0, 0.5, 0.1, 0, 0.2, 0.4), 2, 3), diag(2), 3)$irf
#> [,1] [,2]
#> [1,] 1.000 0
#> [2,] 0.100 0
#> [3,] 0.010 0
#> [4,] 0.001 0
