.- help for kernel programs (STB-26: snp6.1) .- Kernel density estimation ------------------------- kernel_prog varname bandwidth density midpoint where kernel_prog is one of: kernel_prog weight ----------- ------ ^kernsim^ uniform ^kerntria^ triangular ^kernepa^ Epanechnikov ^kernquar^ quartic ^kerntriw^ triweight ^kerncos^ cosinus ^kerngaus^ Gaussian Description ----------- Each of these programs calculates the kernel density estimate for the existing variable ^varname^. The density is estimated at 50 equally-spaced points that cover the range of ^varname^. The ^bandwidth^, a user-specified numeric constant, specifies the width of the density window around each of the 50 points. The 50 estimated densities are stored in the new variable ^density^ and the 50 associated grid points are stored in the new variable ^midpoint^. These procedures can be regarded both as descriptive histogram smoothers as well as nonparametric density estimators. The smoothness of the estimates can be regulated by changing the bandwidth. Wide intervals produce smooth results; narrow intervals result in noiser density estimates. NOTE: These commands are CPU-intensive and may take a long time to execute. Example ------- . ^kernsim snow 10 density midpoint^ . ^list midpoint density in 1/50^ . ^graph density midpoint if _n<=50, c(s) s(p)^ calculates a kernel density estimate of snowfall using the uniform weight function, and lists and graphs the estimate. Authors ------- Isaias H. Salgado-Ugarte, Makoto Shimizu and Toru Taniuchi University of Tokyo, Faculty of Agriculture, Department of Fisheries, Yayoi 1-1-1, Bunkyo-ku Tokyo 113, Japan. FAX: 81-3-3812-0529 EMAIL: isalgado@@tansei.cc.u-tokyo.ac.jp Also see -------- STB: STB-26 snp6.1, STB-16 snp6 On-line: help for @ksm@, @warping@ (if installed)