Department of Scientific Computing    Institute for Numerical Simulation    University of Bonn

Documentation

Download

Programming References

Bug Reports / Suggestions

FAQ

Authors

AdaptiveB

AdaptiveB is the data structure used for the univariate adaptive numerical operations. A common user will have no contact with AdaptiveB since it is used internally only. However, to provide a basic understanding on how AWFD works we give a brief description.

AdaptiveB is designed to achieve a maximum speed for the univariate operations. For each level $l$ from $L_0$ to LMAX (see section 1.2) there is a double array d[] of size $2^{L_0}$ (or $2^{l-1}$ if $l>L_0$ respectively) to store the wavelet coefficients. In the adaptive case not all entries of d are used, but just a subset. For example, if the univariate adaptive index set ${\cal T}$ is

$$\{ (4,0),(4,1),...,(4,16),  (5,0),...,(5,5), (6,0),...,(6,5) \}$$

then on level 4 all entries d[0...16] are used and on level 5 just d[0...5]. The active indices for each level ({0,...,16} for level 4 or {0,...,5} for level 5) are stored in another data structure IndexSet.