Amatdual {adlift} | R Documentation |
Combines filter matrices to produce a refinement matrix A for a wavelet transform.
Amatdual(steps, pointsin, removelist, nbrs, weights, alpha)
steps |
a value indicating which refinement matrix to construct. It refers to the number of points already removed during the transform. |
pointsin |
The indices of gridpoints still to be removed. |
removelist |
a vector of indices into X of the lifted coefficients during the transform (in the order of removal). |
nbrs |
indices of the neighbours used in the last step of the decomposition. |
weights |
the prediction weights obtained from the regression in the prediction step of the transform. |
alpha |
the update weights used to update lengths and coeff. |
The function uses the prediction and update weights to construct the filter matrices Hdual and Gdual. Combining these two matrices results in the refinement matrix Adual.
Adual |
the refinement matrix for the particular step of the transform. |
Hdual |
the high-pass filter matrix for the current step of the transform. |
Gdual |
the low-pass filter matrix for the current step of the transform. |
o |
the indices of nbrs into the vector of pointsin and the steps removed points of the transform. |
alpha |
the update weights used to update lengths and coeff. |
weights |
the prediction weights obtained from the regression in the prediction step of the transform. |
Matt Nunes (matt.nunes@bristol.ac.uk), Marina Popa (Marina.Popa@bristol.ac.uk)
# x<-runif(256) y<-make.signal2("doppler",x=x) a<-fwtnp(x,y,LocalPred=AdaptNeigh,neighbours=2) # A<-Amatdual(90,a$pointsin,a$removelist,a$neighbrs[[90]],a$gamlist[[90]],a$alphalist[[90]]) # A$Adual # #the 90th refinement matrix for the transform above. #