sqrt.est.cv(noisy, wav.filter.number = 4, family = "DaubLeAsymm", thetamin = 0.01, thetamax = 1, thetalength = 11, ti = T)

Performs the SQRT denoising algorithm with an automatic choice of the theta parameter, as described in Section 5 of the paper. Requires WaveThresh.

Takes:

noisy |
The noisy vector; its length has to be a power of two. |

wav.filter.number |
The wavelet filter number of the wavelet basis used for denosing. Can be between 1 and 10 if family = "DaubExPhase", and between 4 and 10 if family = "DaubLeAsymm". |

family |
The wavelet family name of the wavelet basis used for denosing. Can be set to "DaubExPhase" (Daubechies' Extremal Phase) or "DaubLeAsymm" (Daubechies' Least Asymmetric). |

thetamin |
The minimum permitted value of the theta parameter of the thresholding profile; must be between 0 and 1. |

thetamax |
The maximum permitted value of the theta parameter of the thresholding profile; must be between 0 and 1 and no less than thetamin. |

thetalength |
The number of equispaced values of theta, between thetamin and thetamax, which the algorithm examines in order to find the optimum value. Must be a positive integer. |

ti |
If T, then the translation-invariant version of the algorithm is performed. If F, then the algorithm is based on the traditional Discrete Wavelet Transform. |

Returns:

denv |
The denoised vector. |