Quantitative Analysis
Parallel Processing
Numerical Analysis
C++ Multithreading
Python for Excel
Python Utilities
Services
Author

I. Wavelet calculations.
II. Calculation of approximation spaces in one dimension.
1. Calculation of boundary scaling functions.
2. Calculation of boundary wavelets.
3. Testing properties of boundary wavelets and scaling functions.
III. Calculation of approximation spaces in one dimension II.
IV. One dimensional problems.
V. Stochastic optimization in one dimension.
VI. Scalar product in N-dimensions.
VII. Wavelet transform of payoff function in N-dimensions.
VIII. Solving N-dimensional PDEs.
Downloads. Index. Contents.

Calculation of boundary scaling functions.


otivation for calculations of this section was presented in the previous section ( Calculation of approximation spaces ).

We start from the pair MATH introduced in the section ( Symmetric biorthogonal wavelets ) and calculated in the section ( Calculation of scaling functions ). We intend to adapt the procedure of the section ( Adapting GMRA to interval [0,1] ) to have two properties:

1. The procedure should be numerically stable.

2. The Gram matrix MATH should have a low condition number.

The following is the procedure that accomplishes these two goals.

Summary

(Calculation of boundary scaling functions). Let MATH and MATH be the scaling filters calculated as described in the section ( Symmetric biorthogonal wavelets ). Let $\phi$ and $\tilde{\phi}$ be the output of the finite cascade procedure as described in the section ( Calculation of scaling functions ). Fix the scale parameter $d$ .

1. Form the sets MATH as defined in the section ( Adapting GMRA to interval [0,1] ).

2. Calculate MATH and MATH recursively: MATH MATH where the projection is taken in MATH .

3. Calculate MATH and MATH : MATH

4. Form the collection MATH : MATH

5. Calculate MATH and MATH recursively: MATH MATH MATH

6. Form the basis MATH of $V_{d}$ : MATH

The procedure is implemented in the script "bases\phi.py" and tested by the script "_run_bases.py". Experimentation shows that for 6 steps of scaling procedure we produce Gram matrix with condition number below 10 and precision of scaling projection below 0.0001.





Downloads. Index. Contents.


















Copyright 2007