Total number of participants:
Number of Participants not related to CBC:
Number of different nationalities represented:
Total number of speakers: 1
Total number of talks: 1
Abstract
Access to nodal basis functions can be a limiting factor in finite element
codes. Users would like bases for H1, H(div), and H(curl) for any polynomial
degree, but this can be difficult to achieve. Moreover, design decisions are
usually made not by what method is appropriate but by what method is expedient
to implement. I will talk about a general mathematical framework that
translates problems of computing general finite element bases into linear
algebra. As an example, I will talk about the Karper-Mardal-Winther element in
some detail. I will also discuss some of the technological issues, such as
relative merits of using numerical and symbolic methods.