Revisit the trajectory length for nonstationary case

[greenTara]

As part of the parameter printout, let's print out the start and final value of mu that are processed, and also the value of delta. Because the model changes with mu, we should be comparing only parameter sets that have the same range of mu.

[greenTara]

I would like to add a plot of mu (vertical) versus time (horizontal) to the figures where we plot something (usually error, e.g. in the deciles plot #31 ) versus time for the non-stationary case, so we can (visually) match the change in mu to changes in the error.