P2 validation: oracle test and large-scale raw X/Y check

- tests/oracle/test_p2_raw_xy.py: IDAT-assembled raw X/Y and controls
  equal the arrays in oracle GTCs
- validate_p1_large.py: phase-P2 comparison against vendor GTCs with
  diagnostics (channel swap, bead-count disagreements)
- cloud trial: also runs `simurg ingest` on all samples (time/memory),
  disk 250 GB

tests/oracle/test_p2_raw_xy.py

@@ -0,0 +1,36 @@
+"""Phase P2 acceptance on P0 data: raw X/Y and control intensities assembled from IDATs equal the arrays stored
+in oracle GTCs (TOC 1000/1001, 500/501). The same check on vendor GTCs runs in scripts/validate_p1_large.py --p2.
+"""
+
+from __future__ import annotations
+
+import numpy as np
+import pytest
+
+from simurg.formats.bpm import read_bpm
+from simurg.formats.gtc import read_gtc
+from simurg.formats.idat import read_idat
+from simurg.signal.raw_xy import LocusMapCache, assemble_raw_xy, check_coverage
+from tests.conftest import GSA_BPM, GSA_IDATS, ORACLE_GTCS, need
+
+pytestmark = pytest.mark.data
+
+
+def test_raw_xy_equals_oracle_gtcs() -> None:
+    need(GSA_BPM, GSA_IDATS, ORACLE_GTCS)
+    m = read_bpm(GSA_BPM)
+    cache = LocusMapCache(m)
+    gtcs = sorted(ORACLE_GTCS.glob("*.gtc"))
+    assert gtcs
+    for g in gtcs:
+        sid = g.stem
+        chip = sid.split("_")[0]
+        grn = read_idat(GSA_IDATS / chip / f"{sid}_Grn.idat")
+        red = read_idat(GSA_IDATS / chip / f"{sid}_Red.idat")
+        assert check_coverage(grn, m)[0] == 1.0
+        raw = assemble_raw_xy(grn, red, m, cache=cache)
+        ref = read_gtc(g)
+        assert np.array_equal(raw.raw_x, ref.raw_x), sid
+        assert np.array_equal(raw.raw_y, ref.raw_y), sid
+        assert np.array_equal(raw.controls_x, ref.controls_x), sid
+        assert np.array_equal(raw.controls_y, ref.controls_y), sid