cr.sparse.block.bsbl.BSBL_State¶
- class cr.sparse.block.bsbl.BSBL_State(mu_x: jax.Array, r: jax.Array, r_norm_sqr: jax.Array, gammas: jax.Array, Sigma0: jax.Array, lambda_val: float, dmu: float, iterations: int)[source]¶
Sparse Bayesian Learning algorithm state
Attributes
Sigma0
Prior correlation matrices for each block
dmu
Maximum absolute difference between two iterations for means
gammas
Estimated values for gamma for each block
iterations
Number of iterations
lambda_val
Estimated value of the noise variance
mu_x
Mean vectors for each block
r
The residuals
r_norm_sqr
The residual norm squared
x
Solution vector