Ray Tracing in One Weekend
This repository is an implementation of Peter Shirley's book Ray Tracing in One Weekend as implemented in Python 3.
Note on implementation
After getting half way through chapter 5, I realized that performance was going to be an issue with my naive implementation. As a result, I decided to change things. The implementation half way through chapter 5 would take ~21 seconds for a 400 x 200 image. I found a nice blog post which explained and linked to this code, which suggested that an implementation more friendly to numpy arrays would have significant performance benefits. After implementing those changes, the same functionality ran in about 875 ms for a 2000x1000 image. This represents around a 600x speedup, plus it means the implementation should be amenable to CUDA-based numpy acceleration if such a thing exists.