Nonparametric kernel estimation of the distribution function. Bandwidth selection and estimation of related functions Nonparametric kernel distribution function estimation is performed. Three automatic bandwidth selection methods for nonparametric kernel distribution function estimation are implemented: the plug-in of Altman and Leger, the plug-in of Polansky and Baker, and the modified cross-validation of Bowman, Hall and Prvan. The exceedence function, the mean return period and the return level are also computed. Software Depends: date, chron Alejandro Quintela del Rio <aquintela@udc.es> Alejandro Quintela del Rio and Graciela Estevez Perez GPL (>= 2) 2011-03-31 application/tgz http://CRAN.R-project.org/package=kerdiest