Method: deep_prior (600 epochs)
Shown: replication [0] (best)

Priors:
            mean         sd        q05        q95
b0         0.309      0.465     -0.496       1.07
sigma       0.81       0.52      0.179       1.82

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

Loss by target:
       technique       loss
y0     quantiles      0.216
cor  correlation      0.031
y2      identity      0.555

Final loss:
   final_loss
0        2.45
1        5.16
