Discussion Overview
The discussion revolves around the exploration of numerical analysis packages available in C and C++, with participants sharing resources and experiences regarding various libraries such as LAPACK, GSL, and others. The focus includes usability, optimization, and platform compatibility.
Discussion Character
- Exploratory
- Technical explanation
- Debate/contested
Main Points Raised
- One participant inquires about good numerical analysis packages in C/C++, specifically mentioning LAPACK.
- Another participant provides a link to Netlib, suggesting it contains a variety of numerical analysis resources in Fortran, C, and C++.
- A different participant recommends the GNU Scientific Library (GSL) as a good resource for numerical analysis functions, although notes it is primarily for Linux.
- A participant expresses interest in other packages beyond Numerical Recipes in C and questions the optimization of these libraries, particularly cLAPACK.
- One participant shares positive experiences with Netlib packages, mentioning that optimization can vary based on the intended processing type (vector, parallel, scalar).
- A participant mentions LAPACK++ as a C++ version of LAPACK, which is being succeeded by the Template Numerical Toolkit.
- Another participant notes that the optimization of Netlib packages typically exceeds their own implementations.
Areas of Agreement / Disagreement
Participants share resources and experiences, but there is no consensus on which package is definitively the best or most optimized, indicating multiple competing views on usability and performance.
Contextual Notes
Some participants express concerns about platform compatibility, particularly with GSL being Linux-centric, while others highlight varying definitions of optimization in the context of numerical analysis packages.
Who May Find This Useful
Individuals interested in numerical analysis, particularly those working with C/C++ programming and seeking libraries for computational tasks.