Getting AttributeError: 'MultiTrace' object has no attribute 'mean' error

pm.geweke(linear_trace) throws this error

import pymc3 as pm

AttributeError: 'MultiTrace' object has no attribute 'mean'

Here I have pm.summary(linear_trace)

           mean     sd  hdi_3%  hdi_97%  mcse_mean  mcse_sd  ess_mean  ess_sd  ess_bulk  ess_tail  r_hat
Beta0    20.639  0.092  20.547   20.730      0.062    0.052       2.0     2.0       2.0       2.0    inf
Betaa0    0.087  0.189  -0.102    0.275      0.128    0.107       2.0     2.0       2.0       2.0    inf
Betaa1    0.284  0.428  -0.144    0.711      0.291    0.242       2.0     2.0       2.0       2.0    inf
Betaa2    0.517  0.102   0.415    0.619      0.069    0.058       2.0     2.0       2.0       2.0    inf
Betaa3    0.225  0.478  -0.252    0.701      0.324    0.270       2.0     2.0       2.0       2.0    inf
Betaa4    0.788  0.211   0.577    0.999      0.143    0.119       2.0     2.0       2.0       2.0    inf
Betaa5   -0.381  0.201  -0.582   -0.180      0.136    0.114       2.0     2.0       2.0       2.0    inf
Betaa6   -0.258  0.221  -0.478   -0.038      0.150    0.125       2.0     2.0       2.0       2.0    inf
Betaa7    0.652  0.071   0.582    0.722      0.048    0.040       2.0     2.0       2.0       2.0    inf
Betaa8    0.359  0.094   0.266    0.453      0.064    0.053       2.0     2.0       2.0       2.0    inf
Betaa9    0.154  0.198  -0.044    0.351      0.134    0.112       2.0     2.0       2.0       2.0    inf
Betaa10   0.597  0.114   0.483    0.710      0.077    0.064       2.0     2.0       2.0       2.0    inf
Betaa11   0.443  0.537  -0.092    0.979      0.364    0.303       2.0     2.0       2.0       2.0    inf
Betaa12  -0.771  0.079  -0.850   -0.692      0.054    0.045       2.0     2.0       2.0       2.0    inf
Betaa13   0.063  0.061   0.002    0.124      0.041    0.034       2.0     2.0       2.0       2.0    inf
Betaa14   0.087  0.132  -0.044    0.219      0.090    0.075       2.0     2.0       2.0       2.0    inf
Betaa15   0.766  0.242   0.524    1.008      0.164    0.137       2.0     2.0       2.0       2.0    inf
sigma2    0.000  0.000   0.000    0.000      0.000    0.000       2.0     2.0       2.0       2.0    inf

geweke takes a 1D trace of some (single) parameter. So try something like this:

pm.geweke(linear_trace['Beta0'])