Department of Statistics, UCLA

This book is difficult to review, because it is at the same time charmingly interesting and maddeningly insular. In the 1970s the school of Jean-Paul Benz?ecri attained a dominant position in French statistics. This was, to a large extent, because there was a huge vacuum between the probabilists in control of academics and the increasing demands for computerized data analysis. For at least 10 years, French statistics was more or less identical to Analyse des Donn?ees, or, more specifically, to the dynamic duo Analyse des Correspondances and Taxonomie (a.k.a. cluster analysis)