linear binning for univariate data, given the binwidth and optionally the origin of the bin grid. The smallest grid with width d that covers the data is found and the data are binned to this grid using the linear binning rule.

direct computation of the autocovariances of the bincounts needed for fast computation of the kernel estimates of the integrated squared density derivatives.

estimates a univariate density of a random variable that is convoluted with a Gaussian random variable. The estimation is based on deconvolution by inverting the characteristic function.

This is a variant of denxbwcrit and denbwcrit using linear binning for fast computation. All kernel estimates of the integrated squared densities or density derivatives are computed approximately by Rdenbest which uses linear binning.

determines from a range of bandwidths the optimal one using one of the following bandwidth selection criteria: Least Squares Cross Validation (lscv), Biased Cross Validation (bcv), Smoothed Cross Validation (scv), Jones, Marron and Park Cross Validation (jmp), Park and Marron Plug-in (pm), Sheather

automatic bandwidth selection for univariate and multivariate local polynomial regression. It also fits a smooth curve to (x,y) using the obtained local bandwidths.

automatic bandwidth selection for univariate and multivariate local polynomial regression. It also fits a smooth curve to (x,y) using the resulting local bandwidths.

Carries out a penalized functional principal component analysis (PCA) based on the coefficient matrix for functional data. It is possible to choose a smoothing parameter objectively.

computes the Nadaraya-Watson leave-one-out estimator without binning using the quartic kernel. Prior to estimation, looreg sorts the data. The sorted data, along with the sorted leave-one-out regression estimates, are returned as an output.

Estimates the derivative of a regression function (including the 0th derivative) by local polynomial fits on a grid. This quantlet can be used for univariate or multivariate regression estimation.

lprotint computes the integral of the (p+1)st derivative of a polynomial of order (p+3), this function is used to find rule-of-thumb bandwidth for local polynomial regression and derivative estimation

evaluates a kernel estimate of an integrated squared density (derivative) using the normal kernel for a (vector of) bandwidth(s) h. This quantlet is a variation of Rdenxest and uses linearly prebinned data for faster computation.

determines the optimal from a range of bandwidths by one using the resubstitution estimator with one of the following penalty functions: Shibata's penalty function (shi), Generalized Cross Validation (gcv), Akaike's Information Criterion (aic), Finite Prediction Error (fpe), Rice's T function (rice

computes uniform confidence bands with prespecified confidence level for univariate regression using the Nadaraya-Watson estimator. The computation uses WARPing.

computes pointwise confidence intervals with prespecified confidence level for univariate regression using the Nadaraya-Watson estimator. The computation uses WARPing.

determines the optimal from a range of bandwidths by one using the resubstitution estimator with one of the following penalty functions: Shibata's penalty function (shi), Generalized Cross Validation (gcv), Akaike's Information Criterion (aic), Finite Prediction Error (fpe), Rice's T function (rice