neighbayes.models.priors.PanelBasePriors¶

class neighbayes.models.priors.PanelBasePriors(beta_mu=None, beta_sigma=None, sigma2_alpha=2.0, sigma2_beta=None, sigma_sigma=10.0, nu=4.0)[source]¶

Priors shared by every panel model.

beta_mu, beta_sigma: Normal prior on regression coefficients. Both default to None, which means the model resolves a weakly-informative, data-scaled Gelman et al. (2008) prior at construction (see BasePriors for the formula).

sigma2_alpha, sigma2_beta: Inverse-gamma prior on the observation-noise variance \(\sigma^2\). Default alpha=2.0; sigma2_beta defaults to Var(y) when None.

sigma_sigma[source]¶

Half-normal scale on \(\sigma\). Retained for backward compatibility with callers that still pass it; unused by the Gaussian path, which uses sigma2_alpha / sigma2_beta.

Type:¶: float

nu[source]¶

Student-t degrees of freedom used when robust=True. Fixed rather than sampled; see BasePriors.

Type:¶: float

__init__(beta_mu=None, beta_sigma=None, sigma2_alpha=2.0, sigma2_beta=None, sigma_sigma=10.0, nu=4.0)[source]¶

Methods

`__init__`([beta_mu, beta_sigma, ...])

Attributes

`beta_mu`
`beta_sigma`
`nu`
`sigma2_alpha`
`sigma2_beta`
`sigma_sigma`

beta_mu = None[source]¶

beta_sigma = None[source]¶

nu = 4.0[source]¶

sigma2_alpha = 2.0[source]¶

sigma2_beta = None[source]¶

sigma_sigma = 10.0[source]¶