From 5926446846e82d5c155af825bd39a1cf52e036fc Mon Sep 17 00:00:00 2001
From: MasanariHosokawa <masanari.hosokawa@iter.org>
Date: Wed, 9 Jul 2025 08:40:06 +0200
Subject: [PATCH] adding a ruleset file for cross IDS validation

---
 .../equilibrium_and_core_profiles.py          | 89 +++++++++++++++++++
 1 file changed, 89 insertions(+)
 create mode 100644 imas_validator/assets/rulesets/scenarios/equilibrium_and_core_profiles.py

diff --git a/imas_validator/assets/rulesets/scenarios/equilibrium_and_core_profiles.py b/imas_validator/assets/rulesets/scenarios/equilibrium_and_core_profiles.py
new file mode 100644
index 0000000..0cf06ce
--- /dev/null
+++ b/imas_validator/assets/rulesets/scenarios/equilibrium_and_core_profiles.py
@@ -0,0 +1,89 @@
+"""
+Validation rules of ITER scenario database for the IDSes
+``equlibrium`` and ``core_profiles``.
+"""
+
+import numpy as np
+
+
+def get_samples(t1, f1, t2, f2, n=5):
+    """
+    Generate interpolated sample values from two time series for comparison.
+
+    This function takes two time series (t1, f1) and (t2, f2), generates `n`
+    random time points within the union of their time spans, and returns the
+    interpolated values from both series at those sampled time points.
+
+    Parameters
+    ----------
+    t1 : array_like
+        Time array corresponding to the first data series `f1`.
+        Must be 1D and sorted.
+    f1 : array_like
+        Value array corresponding to `t1`.
+    t2 : array_like
+        Time array corresponding to the second data series `f2`.
+        Must be 1D and sorted.
+    f2 : array_like
+        Value array corresponding to `t2`.
+    n : int, optional
+        Number of sample points to generate (default is 5).
+
+    Returns
+    -------
+    tuple of ndarray or bool
+        A tuple (r1, r2), where:
+            - r1 : interpolated values from `f1` at sampled time points
+            - r2 : interpolated values from `f2` at the same time points
+        If inputs are invalid (e.g. empty arrays), returns False.
+
+    Notes
+    -----
+    - Uses linear interpolation (`np.interp`) for both series.
+    - Returns False if either time or value array is empty.
+    - Time sampling is done uniformly over the combined time span of t1 and t2.
+    """
+
+    if len(t1) == 0 or len(t2) == 0:
+        return None
+
+    tmin = min(t1[0], t2[0])
+    tmax = max(t1[-1], t2[-1])
+
+    if abs(tmax - tmin) < 1e-6:
+        ts = [tmin]
+    else:
+        ts = (tmax - tmin) * np.random.random_sample((n,)) + tmin
+
+    if len(f1) > 0 and len(f2) > 0:
+        r1 = np.interp(ts, t1, f1)
+        r2 = np.interp(ts, t2, f2)
+        return (r1, r2)
+    return None
+
+
+@validator("equilibrium:0", "core_profiles:0")
+def validate_inter_signal(eq, cp):
+    """
+    Validate that IDS fields in equilibrium and core_profiles match.
+    """
+
+    # ip check
+    s = get_samples(
+        eq.time_slice.coordinates[0],
+        np.array([ts.global_quantities.ip for ts in eq.time_slice]),
+        cp.global_quantities.ip.coordinates[0],
+        cp.global_quantities.ip,
+    )
+    if s:
+        Approx(s[0], s[1])
+
+    # b0 check
+    s = get_samples(
+        eq.vacuum_toroidal_field.b0.coordinates[0],
+        eq.vacuum_toroidal_field.b0,
+        cp.vacuum_toroidal_field.b0.coordinates[0],
+        cp.vacuum_toroidal_field.b0,
+    )
+    if s:
+        Approx(s[0], s[1])