Vintage material

Some vintage material
possibly still of interest

Non-parametric estimation of the distribution function and quantiles
The paper `A note on the estimation of a distribution function and quantiles by a kernel method' published in Biometrika, vol.68 (1981), pp.326-8, reports summary results on the stated problem without providing the underlying algebraic computations, but stating that `Further details are available from the authors'. People that asked about these `further details' were sent a copy of the unpublished 1979 paper, which includes the algebraic work and other information. Thanks to the current internet facilities, the original paper is made available here.

European Young Statisticians Meeting, 1981
The 1981 Meeting of European Young Statisticians was held in Bressanone/Brixen, Italy. It gathered people from all across Europe, not a trivial achievement in those years, especially in the case of not-yet-established people from certain countries. The book of abstracts and a set of photos taken by G√ľnther Sawitzki represent, presumably, the only available documents of that event. Note that a few names of participants do not show in the book of abstracts; this reflects the communication difficulties of that time, so that not all participants could submit their written contribution in suitable time.

Old Faithful geyser data
The paper by Azzalini and Bowman `A look at some data on the Old Faithful geyser', published in Applied Statistics, vol. 39 (1990), pp. 357-265, has examined data on the geyser eruptions which have occurred in a stretch of 15 consecutive days in August 1985. These data are available here both in the form of original records collected by the park rangers and as a CVS file where the data have been transcribed.

Note 1: these data include some variables not used in the 1990 paper.
Note 2: an earthquake in 1998 has modified the local geological structure; hence the current pattern of eruptions is not the same as in 1985.


Homepage of Adelchi Azzalini
Homepage of `Dipartimento di Scienze Statistiche'