Save stats for z=0

[jtravs]

Is there a reason why we don't save the statistics for z=0? This would be very useful. It would also be nice to save for z=L too.

[jtravs]

I'm guessing it is due to the interpolation, which deals with the exact end cases. Maybe a solution would be to save statistics whenever we save the field, using the interpolated field. And otherwise at the natural grid points when not saving the field?

[chrisbrahms]

There's no intentional reason, no. It happens because the stepfun in the solver is only called the first time after the first step, at which point the "current" field is already propagated.

Maybe a solution would be to save statistics whenever we save the field, using the interpolated field. And otherwise at the natural grid points when not saving the field?

So you mean adding extra stats points, one for each time we save the field? I think this would work, yes.

[jtravs]

OK, I'm testing a PR for this now.