simulate(var::VAR1, N::Int, T::Int, x0::Vector{Float64}
         stochastic::Bool=true, irf::Bool=false, e0::Vector{Float64}=zeros(0))

Construct N simulated paths of length T the vector autoregressive process var, starting each simulated path from x0.

If irf is true, then e0 must be a vector of the same length as var.μ and e0 will be used as the initial shocks.

If stochastic is true, construct a sequence of shocks by drawing N paths of length T from the distribution Normal(0, var.Σ) (where 0 is a vector of zeros of the appropriate length). Otherwise, if stochastic is false, set all shocks to 0 (unless irf is true – see above).

The output is an AxisArray contining variables, paths, and time on the 3 axes