Lect02-2010

# Lect02-2010 - ProjectionanditsImportance...

This preview shows pages 1–9. Sign up to view the full content.

Slide 1 / 39 Projection and its Importance in Scientific Computing ________________________________________________ CS 594 Lecture Notes 01/20/2010 Stan Tomov EECS Department The University of Tennessee, Knoxville January 20, 2010

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide 2 / 39 Contact information     office :  Claxton 330     phone : (865) 974-6317     email  :  tomov@cs.utk.edu Additional reference materials: [1] R.Barrett, M.Berry, T.F.Chan, J.Demmel, J.Donato, J. Dongarra, V. Eijkhout,       R.Pozo, C.Romine, and H.Van der Vorst,  Templates for the Solution of Linear Systems:       Building Blocks for Iterative Methods (2 nd  edition)       http://netlib2.cs.utk.edu/linalg/html_templates/Templates.html [2] Yousef Saad , Iterative methods for sparse linear systems (1st edition)        http://www-users.cs.umn.edu/~saad/books.html
Slide 3 / 39 Topics (as related to  scientific computing )        Projection in Scientific Computing               PDEs, Numerical  solution, Tools, etc.             Sparse matrices,  parallel implementations                     Iterative Methods

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide 4 / 39 Outline Part I – Fundamentals Part II – Projection in Linear Algebra Part III – Projection in Functional Analysis (e.g. PDEs) HPC with Multicore and GPUs
Slide 5 / 39 Part I Fundamentals

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide 6 / 39 What is Projection? Here are two examples        (from linear algebra)                                       (from functional analysis)   u Pu e P : orthogonal projection of        vector u on e 0 1 1 e = 1 Pu u = f(x) P : best approximation (projection) of f(x)      in span{ e }   C[0,1] The error (u – Pu) to be orthogonal to vector e
Slide 7 / 39 Definition Projection  is a linear transformation P from                       a linear space V to itself such that                                P 2  = P equivalently               Let V is direct sum of subspaces V 1  and V 2                                V = V 1    V 2               i.e. for   u   V there are unique u 1  V 1  and u 2 V 2  s.t.                                   u = u 1 +u 2                        Then P: V V 1  is defined for    u   V as  Pu   u 1

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide 8 / 39 Importance in Scientific  Computing To compute approximations Pu   u where   dim V 1  <<  dim V                           V = V 1    V 2 When computation directly in V is not feasible or even possible.
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 04/01/2010 for the course COMPUTER S cs202 taught by Professor Jiuhui during the Spring '08 term at 東京国際大学.

### Page1 / 39

Lect02-2010 - ProjectionanditsImportance...

This preview shows document pages 1 - 9. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online