\begin{table}%t11 
\caption{\label{tab_lensmodel_radio}General results of the best models. }
\small%\centerline
{
\begin{tabular}{ccccc}
\hline\hline
\noalign{\smallskip}
Model & $\chi^2$ Optical & $\chi^2$ Radio & Aver rms & Log(evidence) \\
\noalign{\smallskip}
\hline 
\noalign{\smallskip}
{{\textsc{mfinal}}}  & 11.60 & $-$   & 0.26 & --24.503 \\
{{\textsc{mzfix}}}   & 18.14 & 35.46 & 0.53 & --45.340 \\
{{\textsc{mzfree}}}  & 13.21 & 3.70  & 0.30 & --29.741 \\
\hline
\end{tabular}}
\smallskip
The columns show:
Nickname of the model (Model), sum of the reduced $\chi^2$ of all the
optical/NIR sets of constraints ($\chi^2$ optical), reduced $\chi^2$
for the radio set of constraints ($\chi^2$ radio), average separation
between the observed and predicted position of all the lensed images
used as constraints (aver rms), probability of the model to reproduce
the observations (log(evidence)). 
\end{table}