mcp
mcp copied to clipboard
Turn off autocorrelation
Turning off would just be changing to 0th order:
segments = list(
y ~ 1 + x,
~ 0 + ar(1), # start AR(1)
~ 0 + ar(0) # Stop all AR
)
Implementation-wise, the "0" would just set zeros for the involved parameters, e.g., ar_[i_] = 0
, i.e., leaving it purely to sigma
to model the residuals.