Eric Wieser

Add `LinearMatrix.from_mapping`, and replace `BladeMap` with it

2

For the example application in spacetime algebra, `bm(x)` is at least 10x faster. In theory we could now also write `split` without using `bm` at all: ```python split = LinearMatrix.from_mapping([...

Following the docs example: ```python >>> from clifford import Cl >>> from clifford.cga import CGA >>> g3, blades = Cl(3) >>> locals().update(blades) >>> g3c = CGA(g3) >>> C = g3c.round(3)...

Add a helper for mapping between different algebras which share basis lades

4

TODO: * Tests * Check docs

Split `MultiVector.__init__` into three factory methods

2

This makes the API a little bit easier to explain, and removes the opportunity for passing orthogonal options at the same time. Let's see whether the resulting docs look ok

Improve performance of multivector arithmetic by avoiding copies

4

Multiplications become roughly 20% faster, and addition 33% faster. There's now a new `copy` argument to the `MultiVector` constructor. Some timings: **Edit:** These timings are out of date, and include...

Today, these are: ``` 198.76s call clifford/test/test_algebra_initialisation.py::TestInitialisation::test_sparse_multiply[conformalize(Cl(3))] 40.55s call clifford/test/test_algebra_initialisation.py::TestInitialisation::test_sparse_multiply[Cl(4)] 36.41s call clifford/test/test_g3c_tools.py::TestRotorEstimation::test_estimate_rotor_optimisation[random_point_pair] 33.02s call clifford/test/test_g3c_tools.py::TestModelMatching::test_fingerprint_match 23.04s call clifford/test/test_g3c_tools.py::TestGeneralLogarithm::test_general_logarithm_conformal[random_point_pair] 20.31s call clifford/test/test_g3c_tools.py::TestRotorEstimation::test_estimate_rotor_sequential[random_sphere] 17.68s call clifford/test/test_g3c_tools.py::TestGeneralLogarithm::test_general_logarithm_conformal[random_line] 15.12s call clifford/test/test_g3c_tools.py::TestGeneralLogarithm::test_general_logarithm_conformal[random_circle] 14.87s call...

Star imports of algebras confuse flake8

1

`from clifford.g3c import *` confuses flake8 with F403 and F405 errors. I don't know if we can do anything about this, but figured I'd open an issue to track it.

Allow running without numba installed?

2

It seems we can't run on pyiodide with numba around: https://github.com/pygae/clifford/issues/96#issuecomment-555491381

Punted from #225, which seemed to still try to run 3.7 on mac. For now, we simply do not run CI on this platform at all.

Tests are missing for the following methods in `MultiVector` - [x] `.__bool__()` (#204) - [ ] `.round()` - [x] `.pseudoScalar` / `I` (#149) - [x] `.blades_list` (#257) - [ ]...