Lecture 6 of 'Scientific Computing' (wi4201)
previous lecture,
next lecture
Collegerama
- The following subjects are discussed:
- Partial pivotting, complete pivotting
- Solve PA = LU and PA u = Pf
- Example partial pivotting
- When is partial pivotting needed?
- A is SPD, Cholesky decomposition
- Partial pivotting is not needed if A is SPD
- Band matrices, sparse matrices
- Unstructured grid, profile method
- Iterative refinement
- Double precision, single precision
- Basic Iterative Methods (BIM)
- error and residual
- splitting of matrix A
- convergence, error iteration, iteration matrix
- Richardson method, spectral radius
- spectral radius less than 1 for fast convergence
- Material is described in pages 11 and 12, 51 - 64 of the lecture notes.
Recommended exercises: 4.13.5 - 4.13.7 and 5.9.1 - 5.9.3
Back to
Lectures of Scientific Computing page.