Prediction interval for smooth

[martin-gmx]

I was stumbling across the question if one could plot for a smooth not only the CI of the moving average but also the prediction interval to cover variability.

Apparently, this is available in the msir package - but it might be a nice addition to xgxr, to be available to me by default.
Details at stackexchange
(https://stats.stackexchange.com/questions/141552/how-to-calculate-prediction-intervals-for-loess)

Thanks, Martin