Closed mohanwu closed 2 years ago
Both bridge and bootstrap methods will use the normal approximation as discussed: log y_t ~ N( log x_t, 1/quad)
Sampling model is still the unapproximated model, ie y_t ~ N( x_t, tau^2)
Commented out 64-bit config in test
Both bridge and bootstrap methods will use the normal approximation as discussed: log y_t ~ N( log x_t, 1/quad)
Sampling model is still the unapproximated model, ie y_t ~ N( x_t, tau^2)
Commented out 64-bit config in test