- introduction to testing with a focus on numerics
- guide to profiling
- introduction to packaging and distribution
- which tool to use for which job (cover core packages)
- data visualization options
- write a numpy ufunc in Python, Cython, and C
- roundup of high-performance options (C, Cython, Numba, Parakeet, etc.)
Thanks to Rob Story for some suggestions. If you’ve got ideas for talks you’d like to see, leave a comment!
(I will be at SciPy 2015, but I’m organizing a Software Carpentry tutorial so I probably won’t be submitting a talk proposal.)
P.S. If you’re a first-time speaker from an underrepresented group thinking about giving a talk at SciPy 2015, I’m offering to help with proposal editing and talk prep.