All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
Massive performance improvements. On i7-9750H CPU (Coffee Lake):
- 20x20 board, photon propagation: 22-34ms -> 0.02ms
- 100x100 board, photon propagation: 741 ms -> 0.02 ms
- 20x20 board, 6 operations: 80 ms -> 1.5ms
- Some performance benchmarks and tests
- Hard-coded photon propagation (a single map, instead of an operator)
- Catching photon operations (rather than recalculating them each step)
- Smarter photon operations (only looking at the difference), i.e.
- This CHANGELOG!
Operator.permuteDimsOutand ``Operator.permuteDimsIn` methods.
- Fixed a problem with a dimension check on permuting non-square tensors.
- Set phase for LR basis (for polarization) and y-spin basis (for spin) so that the singlet state phase stays the same