Generates the translation invariant estimate of x with automatic hardthresholding. It is well-known that nonlinear wavelet estimators are not translation-invariant: if we shift the underlying data set by a small amount, apply nonlinear thresholding and shift the estimator back, then we usually obta
Stein computes the optimal threshold for a vector of data plus noise so that the mean squared error is minimized. Stein uses Stein's unbiased risk estimator for the risk. The quantlet sure uses stein to threshold the father and mother wavelet coefficients.
Sure denoises wavelet coefficients so that the mean squared error is minimized. MSE is estimated by Stein's unbiased risk estimator based on the variance of the coefficients. Sure computes the optimal threshold for the father wavelets and each level of mother wavelets. The input arrays can be obtai
Sure denoises wavelet coefficients. If the stein procedure is chosen, the mean squared error is minimized. MSE is estimated by Stein's unbiased risk estimator based on the variance of the coefficients. Sure computes then the optimal threshold for the father wavelets and each level of mother wavelet
Generates smoothed mother wavelet coeffients and the resulting estimate. x contains the vector of data and l gives the number of father wavelet coefficient (power of 2). h is the vector of wavelet basis coeffients (automatically generated by calling the quantlib wavelet). s contains the threshold v