Perform the Wild Scale-Enhanced (WiSE) bootstrap. Specifically, the user may supply a single or multiple equally-spaced time series and use the WiSE bootstrap to select a wavelet-smoothed model. Conversely, a pre-selected smooth level may also be specified for the time series. Quantities such as the bootstrap sample of wavelet coefficients, smoothed bootstrap samples, and specific hypothesis testing and confidence region results of the wavelet coefficients may be obtained. Additional functions are available to the user which help format the time series before analysis. This methodology is recommended to aid in model selection and signal extraction. Note: This package specifically uses wavelet bases in the WiSE bootstrap methodology, but the theoretical construct is much more versatile.
|Author||Megan Heyman, Snigdhansu Chatterjee|
|Date of publication||2016-04-03 16:55:59|
|Maintainer||Megan Heyman <email@example.com>|
CM20N20S150W: AIRS, IPSL, and MIROC5 Data at 150W
CM20N20S60E: AIRS, IPSL, and MIROC5 Data at 60E
deSeasonalize: De-seasonalize daily, monthly, or data series with IDs
padMatrix: Increase data length to the closest power of 2.
padVector: Increase data length to the closest power of 2.
retrieveBootstrapSample: Construct the bootstrap data series from wavelet coefficients
SimulatedSmoothSeries: Simulated Wavelet-Smoothed Series
SimulatedSNR15Series: Simulated Wavelet Signals with SNR=15
SimulatedSNR25Series: Simulated Wavelet Signals with SNR=25
SimulatedSNR5Series: Simulated Wavelet Series with SNR=5
SimulatedSNR9Series: Simulated Wavelet Signals with SNR=9
smoothTimeSeries: Threshold Wavelet Coefficients to Create Smooth Time Series
WiSEBoot: Wild Scale-Enhanced (WiSE) Bootstrap for Model Selection
WiSEBoot-package: Wild Scale-Enhanced (WiSE) Bootstrap
WiSEConfidenceRegion: WiSE Wavelet Coefficients: Linear Confidence Region
WiSEHypothesisTest: WiSE Wavelet Coefficients: Linear Hypothesis Test