maybe this works for staterror

Author: kratsg

src/pyhf/modifiers/staterror.py

@@ -6,22 +6,25 @@
 from pyhf.exceptions import InvalidModifier
 from pyhf.parameters import ParamViewer
 from pyhf.tensor.manager import get_backend
+from typing import Optional
 
 log = logging.getLogger(__name__)
 
 
-def required_parset(sigmas, fixed: List[bool]):
+def required_parset(sigmas, fixed: List[bool], constraint: Optional[str] = "gaussian"):
     n_parameters = len(sigmas)
     return {
-        'paramset_type': 'constrained_by_normal',
+        'paramset_type': 'constrained_by_normal'
+        if constraint == "gaussian"
+        else 'constrained_by_poisson',
         'n_parameters': n_parameters,
         'is_shared': True,
         'is_scalar': False,
         'inits': (1.0,) * n_parameters,
         'bounds': ((1e-10, 10.0),) * n_parameters,
         'fixed': tuple(fixed),
         'auxdata': (1.0,) * n_parameters,
-        'sigmas': tuple(sigmas),
+        'sigmas' if constraint == "gaussian" else 'factors': tuple(sigmas),
     }
 
 
@@ -36,11 +39,12 @@ def __init__(self, config):
     def collect(self, thismod, nom):
         uncrt = thismod['data'] if thismod else [0.0] * len(nom)
         mask = [True if thismod else False] * len(nom)
-        return {'mask': mask, 'nom_data': nom, 'uncrt': uncrt}
+        constraint = thismod.get('constraint', 'gaussian') if thismod else 'gaussian'
+        return {'mask': mask, 'nom_data': nom, 'uncrt': uncrt, 'constraint': constraint}
 
     def append(self, key, channel, sample, thismod, defined_samp):
         self.builder_data.setdefault(key, {}).setdefault(sample, {}).setdefault(
-            'data', {'uncrt': [], 'nom_data': [], 'mask': []}
+            'data', {'uncrt': [], 'nom_data': [], 'mask': [], 'constraint': []}
         )
         nom = (
             defined_samp['data']
@@ -51,6 +55,9 @@ def append(self, key, channel, sample, thismod, defined_samp):
         self.builder_data[key][sample]['data']['mask'].append(moddata['mask'])
         self.builder_data[key][sample]['data']['uncrt'].append(moddata['uncrt'])
         self.builder_data[key][sample]['data']['nom_data'].append(moddata['nom_data'])
+        self.builder_data[key][sample]['data']['constraint'].append(
+            moddata['constraint']
+        )
 
     def finalize(self):
         default_backend = pyhf.default_backend
@@ -115,12 +122,22 @@ def finalize(self):
 
             for modifier_data in self.builder_data[modname].values():
                 modifier_data['data']['mask'] = masks[modname]
+
             sigmas = relerrs[masks[modname]]
             # list of bools, consistent with other modifiers (no numpy.bool_)
             fixed = default_backend.tolist(sigmas == 0)
             # ensures non-Nan constraint term, but in a future PR we need to remove constraints for these
             sigmas[fixed] = 1.0
-            self.required_parsets.setdefault(parname, [required_parset(sigmas, fixed)])
+
+            constraint = [
+                i
+                for i, v in zip(modifier_data['data']['constraint'], masks[modname])
+                if v
+            ]
+            assert all(constraint[0] == element for element in constraint)
+            self.required_parsets.setdefault(
+                parname, [required_parset(sigmas, fixed, constraint)]
+            )
         return self.builder_data
 
 

src/pyhf/pdf.py

@@ -35,6 +35,7 @@ def _finalize_parameters_specs(user_parameters, _paramsets_requirements):
                 f"Multiple parameter configurations for {parameter['name']} were found."
             )
         _paramsets_user_configs[parameter.get('name')] = parameter
+
     _reqs = reduce_paramsets_requirements(
         _paramsets_requirements, _paramsets_user_configs
     )

src/pyhf/schema/loader.py

@@ -39,3 +39,6 @@ def load_schema(schema_id: str):
             schema = json.load(json_schema)
             variables.SCHEMA_CACHE[schema['$id']] = schema
         return variables.SCHEMA_CACHE[schema['$id']]
+
+
+load_schema(f'{variables.SCHEMA_VERSION}/defs.json')