Abstract
A method based on the Fourier transform theory is proposed to determine the probability density function from a given set of random samples. First the Fourier transform of the probability density function is obtained from the set. If the transform has a long tail, it is to be smoothed. Then an appropriate window is applied to the transform. The inverse of the resulting transform yields the probability density function. The present method is applied to several examples and is shown to be very effective. (c) 1995 The American Physical Society

This publication has 7 references indexed in Scilit: