Method: parametric_prior (600 epochs)
Shown: replication [0, 1] (all)

Hyperparameters:
        rep   estimate     rep_sd        min        max
mu0       0      0.315      0.206     0.0234      0.315
mu0       1     0.0234      0.206     0.0234      0.315
sigma0    0      0.952      0.192      0.681      0.952
sigma0    1      0.681      0.192      0.681      0.952

Priors:
       rep       mean         sd        q05        q95
b0       0      0.309      0.465     -0.496       1.07
b0       1     0.0656       0.87      -1.32       1.66
sigma    0       0.81       0.52      0.179       1.82
sigma    1       0.52       0.31       0.05        1.1

Fit to expert:
        rep     expert   elicited      error
y0[0]     0         -1     -0.916      0.084
y0[0]     1         -1     -0.326      0.674
y0[1]     0          0      0.113      0.113
y0[1]     1          0      0.694      0.694
cor[0]    0       0.01     -0.002     -0.012
cor[0]    1       0.01      0.004     -0.006
y2: draws, not statistics; see its loss in 'Loss by target'.

Loss by target:
     rep    technique       loss
y0     0    quantiles      0.216
y0     1    quantiles        0.9
cor    0  correlation      0.031
cor    1  correlation        0.2
y2     0     identity      0.555
y2     1     identity        1.4

Final loss:
   final_loss
0        2.45
1        5.16

rep_sd, min and max are over all replications.
