We explore a cross-validation approach for threshold selection in wavelet shrinkage and provide R-Package CVThresh for implementation. For details, refer to CVThresh: R Package for Level-Dependent Cross-Validation Thresholding.

CVThresh 1.0.1 depends on non-CRAN version of WaveThresh4 BETA and CVThresh 1.1.0 uses CRAN version of WaveThresh. For CVThresh 1.0.1, package source and pre-complied windows binary are provided. To install windows binary, use RGui menu "Package/Install package(s) from local zip files..." or R command "install.packages". See CRAN for CVThresh 1.1.0.

See version 1.0.1 or version 1.1.0 reference manual for usage and syntax of CVTresh for implementing level-dependent cross-validation for wavelet shrinkage.

Assume that WaveThresh of Nason (1998) and EbayesThresh of Johnstone & Silverman (2005) are pre-installed.

We provide 7 examples. To carry out all examples in simulation studies and real examples, follows Introduction. One may download example R code or copy & paste R code into the R GUI.

• Indexing example according to various cross-validation scheme
• Imputation example by Wavelet
• Analysis of IPD by level-dependent cross-validation thresholding
• Plot of four test functions
• Simulation code under various setting
• Flexibility of proposed level-dependent cross-validation. (Before running the code, download Result of one-dimensional cross validation and Result of two-dimensional cross validation)
• Reconstruction example of noisy Lennon Image