Abstract
In the Abel inversion of a function whose values are determined experimentally on a discrete set of points, random errors are amplified considerably. This paper presents a procedure to smooth the data before inversion, and thus keep the error amplification under control.
Keywords