| Library: | cafpe |
| See also: | tp/cafpe/hoptest |
| Quantlet: | hsilv | |
| Description: | Quantlet to compute Silverman's rule-of-thumb bandwidth for density estimation using either the gaussian or uniform kernel |
| Usage: | h = hsilv(sigy,d,ntotal,kerntype) | |
| Input: | ||
| sigy | scalar, standard deviation of regressors | |
| d | number of continuous regressors | |
| ntotal | number of observations for estimation | |
| kerntype | type of kernel used: "gaussian" or "quartic" | |
| Output: | ||
| h | Silverman's rule-of-thumb bandwidth | |
pathcafpe = "tp/cafpe/" ; path of CAFPE quantlets func(pathcafpe + "hsilv.xpl") ; load required XploRe files of CAFPE d = 2 ntotal = 100 kerntype = "gaussian" y = normal(ntotal,d) sigy = prod(var(y)')^(1/d) h = hsilv(sigy,d,ntotal,kerntype) "h" h
returns Silverman's rule-of-thumb bandwidth
| Library: | cafpe |
| See also: | tp/cafpe/hoptest |