Kernel smoothing is one of the most widely used nonparametric data smoothing techniques. We introduce a new R package, Disake, for computing discrete associated kernel estimators for probability mass function. When working with a kernel estimator, two choices must be made: the kernel function and the smoothing parameter. The Disake package focuses on discrete associated kernels and also on cross-validation and local Bayesian techniques to select the appropriate bandwidth. Applications on simulated data and real data show that the binomial kernel is appropriate for small or moderate count data while the empirical estimator or the discrete triangular kernel is indicated for large samples.
This article deals with the localization of eigenvalues of a large sparse and not necessarilysymmetric matrix in a domain of the complex plane. It combines two studies carried out earlier.The first work deals with the effect of applying small perturbations on a matrix, and referred to ase -spectrum or pseudospectrum. The second study describes a procedure for counting the numberof eigenvalues of a matrix in a region of the complex plain surrounded by a closed curve. The twomethods are combined in order to share the LU factorization of the resolvent, that intervenes in thetwo methods, so as to reduce the cost. The codes obtained are parallelized.
