Regression tests

End-to-end tests compare jitr against committed outputs from Frescox and TALYS. Reference CSVs are committed; neither external code is required in CI.

Running

uv run pytest tests/regression/

Frescox cases (F1–F10)

Ten cases against LLNL Frescox: eight elastic for p/n + 78Ni, and two quasi-elastic (p,n) to the isobaric analog state of 48Ca.

Case	Reaction	Projectile	E_lab (MeV)	Source deck
F1	elastic	p	6.9	B1-example-el.out (block 1)
F2	elastic	p	11.0	B1-example-el.out (block 2)
F3	elastic	p	49.35	B1-example-el.out (block 3)
F4	elastic	p	100	B1-high-el.in (block 1)
F5	elastic	p	200	B1-high-el.in (block 2)
F6	elastic	n	49.35	B1_n-high-el.in (block 1)
F7	elastic	n	100	B1_n-high-el.in (block 2)
F8	elastic	n	200	B1_n-high-el.in (block 3)
F9	48Ca(p,n)48Sc(IAS)	p	25	Ca48_pn_IAS_25MeV.in
F10	48Ca(p,n)48Sc(IAS)	p	35	Ca48_pn_IAS_35MeV.in

F9 and F10 exercise jitr.xs.quasielastic_pn rather than jitr.xs.elastic. Frescox does its own channel-spin algebra, so they are an independent check on the spin-flip terms of the (p,n) amplitude, which the elastic cases cannot see. jitR reproduces both to better than 1.3e-4 at every angle.

TALYS cases (T1–T4)

Four JLM/JLMB elastic-scattering cases against TALYS 2.2 for 120Sn at 10 MeV:

Case	Projectile	jlmmode	Notes
T1	n	0	standard JLMB normalization
T2	n	2	stronger JLMB normalization
T3	p	0	Coulomb code path
T4	p	2	Coulomb + stronger normalization

Harness details

• Regression harness
• Frescox regression references
• TALYS regression references