add test for sum_rules=TSR

n3fit/src/n3fit/layers/msr_normalization.py

@@ -105,7 +105,10 @@ def __init__(self, mode: str = "ALL", replica_seeds=None, **kwargs):
         else:
             raise ValueError(f"Mode {mode} not accepted for sum rules")
 
-        self._replicas = len(replica_seeds)
+        if replica_seeds is None:
+            self._replicas = 1
+        else:
+            self._replicas = len(replica_seeds)
 
         indices = []
         self.divisor_indices = []
@@ -119,6 +122,9 @@ def __init__(self, mode: str = "ALL", replica_seeds=None, **kwargs):
                 [np.repeat(VSR_CONSTANTS[c], self._replicas) for c in VSR_COMPONENTS]
             )
         if self._tsr_enabled:
+            if replica_seeds is None:
+                raise ValueError("To use sum_rules=TSR a list of seeds must be provided")
+
             self.divisor_indices += [IDX[TSR_DENOMINATORS[c]] for c in TSR_COMPONENTS]
             indices += [IDX[c] for c in TSR_COMPONENTS]
 

n3fit/src/n3fit/tests/regressions/quickcard.yml

@@ -29,9 +29,9 @@ theory:
 
 ############################################################
 genrep: True    # on = generate MC replicas, False = use real data
-trvlseed: 5
+trvlseed: 3
+nnseed: 2
 mcseed: 1
-nnseed: 20
 
 load: "weights.weights.h5"
 

n3fit/src/n3fit/tests/test_msr.py

@@ -1,4 +1,5 @@
 import numpy as np
+import pytest
 
 from n3fit.backends import operations as op
 from n3fit.layers import MSR_Normalization
@@ -76,3 +77,18 @@ def test_vsr():
         ]
     )
     np.testing.assert_allclose(output, known_output, rtol=1e-5)
+
+
+def test_tsr():
+    """Test the sum rules used in polarized fits"""
+    with pytest.raises(ValueError):
+        # Check that seeds _are_ needed
+        layer = MSR_Normalization(mode='TSR')
+
+    layer = MSR_Normalization(mode="TSR", replica_seeds=[3])
+    output = apply_layer_to_fixed_input(layer)
+    # They should _all_ be 1.0 except for entries 9 and 10
+    known_output = np.ones((1, 1, 14))
+    known_output[0, 0, 9] = 1.1133982
+    known_output[0, 0, 10] = -0.9901034
+    np.testing.assert_allclose(output, known_output, rtol=1e-5)