FEAT: implement cached.unfold for amplitude models

[redeboer]

Implemented a convenience function, cached.unfold(), which does two things:

1. It combines and 'unfolds' (evaluates) the model.intensity and its model.amplitudes in one go.
2. It unfolds and combines a top expression (like model.intensity) and its definitions (like model.amplitudes) efficiently by first calling doit on the top expression, then calling doit on all the expressions in the definitions, and finally combining them with xreplace. This is more efficient than using xreplace and then doit on the whole result. The function also uses caching internally.

Note that cached.unfold takes either a single amplitude model or a top expression and a list of definitions. This is motivated by ComPWA/polarimetry, where instead of using model.intensity as the top expression, you use the polarimeter field definition $\vec\alpha$.

See also this test

ampform/tests/sympy/test_cached.py

Lines 44 to 47 in 8bb9c07

	amplitudes = {k: v.doit() for k, v in model.amplitudes.items()}
	intensity_expr_direct = model.intensity.doit().xreplace(amplitudes)
	intensity_expr_unfold = cached.unfold(model.intensity, model.amplitudes)
	assert intensity_expr_direct == intensity_expr_unfold

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB