Functions
	run (mode)

Variables
int	I = 1000

	DOSE = float(os.environ.get("PDOSE", "20"))

list	KEYS

dict	out = {}

	w

	o

	res

	a

	b

tuple	rel = (b - a) / a * 100.0 if a else float("nan")

Detailed Description

Is d_111's residual a CLOSURE error, or a PATH difference?

CLAUDE.md S1: with I_bin = 0 and i_discrete = I every equation is discrete, so
bin_moment reduces to full_CD.  Run that against the real discrete arm on an
identical grid at the dose where the problem shows:

    agree    -> the two paths solve the same equations; the -5% on d_111 is
                genuine closure error and must be explained by the closure
    disagree -> a path difference, which would ALSO explain why the error does
                not refine away with bin count

check_bm_vs_discrete.py already asks this, but at 1e-2 dpa.  The M2R deviation
is measured at 20 dpa, three orders of magnitude further out, so a difference
that only opens up as the distribution broadens is invisible to it.

Function Documentation

◆ run()

check_path_equivalence_20dpa.run ( mode )

Definition at line 28 of file check_path_equivalence_20dpa.py.

Variable Documentation

◆ a

check_path_equivalence_20dpa.a

Definition at line 66 of file check_path_equivalence_20dpa.py.

◆ b

check_path_equivalence_20dpa.b

Definition at line 66 of file check_path_equivalence_20dpa.py.

◆ DOSE