hypChan2020()

Populate a hyperparamater structure for models based on Chan (2020) priors

Arguments

    c1: hyperparameter on own lags 
    c2: hyperparameter on other lags
    c3: hyperparameter on the constant
    ρ:  ar(1) parameter for the stochastic volatility (Chan2020csv only)
    σ_h2: variance of the log-volatility (Chan2020csv only)

hypBGR2010()

Generate a structure with hyperparameters for Banbura, Giannone, and Reichlin (2010) Large Bayesian VARs

Arguments

lambda: shrinkage parameter between AR(1) model and maximum likelihood. Default 0.1
epsi:

Examples

• Using default values. Note that the main function will auto-generate this for you. If you don't plan to change any there is rarely need to ever call it.

julia> hyp = hypBGR2010()
hypBGR2010
  lambda: Float64 0.1
  epsi: Float64 0.001

• If a tighter prior (shrinkage towards AR(1)) is desired due to a larger VAR:

julia> hyp = hypBGR2010(lambda=0.05)
hypBGR2010
  lambda: Float64 0.05
  epsi: Float64 0.001