Data smoothing is too large and vague expression.
Rather, understand "Data Rectification"

Too many of this gadgets and lot missing. One very
useful and powerful is the "Gaussian Ksmooth". For
this particular example, there is one even better,
but will do NO better in general applications.
Matlab muffed this data set [splinesmooth].
Past the rectification, if you want an analytical
function you will have to attempt a Padé rational
fraction, complete success for this demo.

Jean

Smooth Ksmooth.sm (21 KiB) downloaded 64 time(s).