analyze_ir.varlse {bvhar}

Impulse Response Analysis

Description

Computes responses to impulses or orthogonal impulses

Usage

## S3 method for class 'varlse'
analyze_ir(
  object,
  lag_max = 10,
  orthogonal = TRUE,
  impulse_var,
  response_var,
  ...
)

## S3 method for class 'vharlse'
analyze_ir(
  object,
  lag_max = 10,
  orthogonal = TRUE,
  impulse_var,
  response_var,
  ...
)

## S3 method for class 'bvharirf'
print(x, digits = max(3L, getOption("digits") - 3L), ...)

analyze_ir(object, lag_max, orthogonal, impulse_var, response_var, ...)

## S3 method for class 'bvharirf'
knit_print(x, ...)

Arguments

object

Model object

lag_max

Maximum lag to investigate the impulse responses (By default, 10)

orthogonal

Orthogonal impulses (TRUE) or just impulses (FALSE)

impulse_var

Impulse variables character vector. If not specified, use every variable.

response_var

Response variables character vector. If not specified, use every variable.

...

not used

x

bvharirf object

digits

digit option to print

Value

bvharirf class

Responses to forecast errors

If orthogonal = FALSE, the function gives W_j VMA representation of the process such that

Y_t = \sum_{j = 0}^\infty W_j \epsilon_{t - j}

Responses to orthogonal impulses

If orthogonal = TRUE, it gives orthogonalized VMA representation

\Theta

. Based on variance decomposition (Cholesky decomposition)

\Sigma = P P^T

where P is lower triangular matrix, impulse response analysis if performed under MA representation

y_t = \sum_{i = 0}^\infty \Theta_i v_{t - i}

Here,

\Theta_i = W_i P

and v_t = P^{-1} \epsilon_t are orthogonal.

References

Lütkepohl, H. (2007). New Introduction to Multiple Time Series Analysis. Springer Publishing.

See Also

VARtoVMA()

VHARtoVMA()