NormalGammaPrior._tracked_label

NormalGammaPrior._tracked_label()

What the averaged diagnostic is.

The result’s shrinkage_label; it contains "scale", which is how shrinkage_scales() knows the diagnostic is a scale rather than an inclusion probability. The same label as the horseshoe’s and the Dirichlet-Laplace’s, because all three report the prior standard deviation of the standardized coefficient.

Return type:

str

Returns:

The label.

Example

>>> NormalGammaPrior()._tracked_label()
'posterior mean local-global scale of the standardized coefficient'