Merge pull request #133 from Climate-REF/fix-pycmec-attributes

fix the validation error when 'attributes' value is a dict

Author: minxu74

changelog/133.fix.md

@@ -0,0 +1 @@
+Fixed the validation error when 'attributes' value is a dict

packages/ref-core/src/cmip_ref_core/pycmec/metric.py

@@ -157,18 +157,24 @@ def _check_nested_dict_keys(cls, nested: dict[Any, Any], metdims: dict[Any, Any]
             raise ValueError("Error in dicts of Results")
 
         for key, value in nested.items():
-            if isinstance(value, dict) and level < len(metdims[MetricCV.JSON_STRUCTURE.value]) - 1:
+            if key == MetricCV.ATTRIBUTES.value:
+                continue
+
+            elif isinstance(value, dict) and level < len(metdims[MetricCV.JSON_STRUCTURE.value]) - 1:
                 cls._check_nested_dict_keys(value, metdims, level + 1)
             elif isinstance(value, dict):
-                StrNumDict(value)
+                tmp = dict(value)
+                if MetricCV.ATTRIBUTES.value in tmp:
+                    tmp.pop(MetricCV.ATTRIBUTES.value)
+                StrNumDict(tmp)
 
     @field_validator("root", mode="after")
     @classmethod
     def _validate_results(cls, rlt: Any, info: ValidationInfo) -> Any:
         """Validate a MetricResults object"""
         if not isinstance(info.context, MetricDimensions):
             s = "\nTo validate MetricResults object, MetricDimensions is needed,\n"
-            s += "please use model_validate(Results, context=MetricDimensions to instantiate\n"
+            s += "please use model_validate(Results, context=MetricDimensions) to instantiate\n"
             raise ValueError(s)
         else:
             # results = rlt.root
@@ -295,7 +301,11 @@ def _fill(cls, mdict: dict[Any, Any], mdims: dict[Any, Any], level: int = 0) ->
                 mdict[key] = {}
 
         for key, value in mdict.items():
-            if isinstance(value, dict) and level < len(mdims[MetricCV.JSON_STRUCTURE.value]) - 1:
+            if (
+                isinstance(value, dict)
+                and level < len(mdims[MetricCV.JSON_STRUCTURE.value]) - 1
+                and key != MetricCV.ATTRIBUTES.value
+            ):
                 cls._fill(value, mdims, level + 1)
 
     @classmethod
@@ -321,8 +331,8 @@ def merge(cls, metric_obj1: Any, metric_obj2: Any) -> Self:
 
         merged_obj_dims = MetricDimensions.merge_dimension(mobj1.DIMENSIONS, mobj2.DIMENSIONS)
 
-        result1 = mobj2.RESULTS
-        result2 = mobj1.RESULTS
+        result1 = mobj1.RESULTS
+        result2 = mobj2.RESULTS
         merged_obj_rlts = cls._merge(dict(result1), result2)
 
         cls._fill(merged_obj_rlts, merged_obj_dims.root)

packages/ref-core/tests/unit/pycmec/cmec_testdata/cmec_metric_sample.json

@@ -33,6 +33,13 @@
             "Hydrology Cycle": {
                 "overall score": 0.26,
                 "bias": 0.70
+            },
+            "attributes":{
+                "resolution": "low",
+                "attr0": "1980-2015",
+                "attr1":{},
+                "attr2":[],
+                "attr3": -999
             }
         },
         "CESM2": {
@@ -43,18 +50,44 @@
             "Hydrology Cycle": {
                 "overall score": 0.61,
                 "bias": 0.18
+            },
+            "attributes":{
+                "resolution": "low",
+                "attr0": "1980-2015",
+                "attr1":{},
+                "attr2":[],
+                "attr3": -999
             }
         },
         "IPSL-CM5A-LR": {
             "Ecosystem and Carbon Cycle": {
                 "overall score": 0.08,
                 "bias": 0.92,
-                "rmse": 0.34
+                "rmse": 0.34,
+                "attributes":{
+                    "overall score": "ILAMB scoring system",
+                    "bias": "mean bias",
+                    "rmse": "root mean squre error"
+                }
             },
             "Hydrology Cycle": {
                 "overall score": 0.67,
                 "rmse": 0.68
+            },
+            "attributes":{
+                "attr1":{
+                    "attr2":{
+                        "attr3":{}
+                    }
+                }
             }
+        },
+        "attributes":{
+            "attr0": "1980-2015",
+            "attr1":{},
+            "attr2":[],
+            "attr3": -999,
+            "attr4":"first level"
         }
     }
 }

packages/ref-core/tests/unit/pycmec/test_cmec_metric.py

@@ -198,18 +198,50 @@ def test_metric_merge():
             "GFDL-ESM2M": {
                 "NinoSstDiversity_2": -75,
                 "BiasTauxLonRmse": 5.959564210237568,
-            }
+                "attributes": {
+                    "NinoSstDiversity_2": "Nino SST diversity",
+                    "BiasTauxLonRmse": "Bias Taux RMSE",
+                },
+            },
+            "attributes": {
+                "package": "pmp",
+                "NinoSstDiversity_2": {},
+            },
         },
     }
     dict_ilamb = {
         "DIMENSIONS": {
             "json_structure": ["model", "metric"],
-            "model": {"E3SM": {"name": "E3SM"}, "CESM": {"name": "CESM"}},
+            "model": {
+                "E3SM": {"name": "E3SM"},
+                "CESM": {"name": "CESM"},
+                "GFDL-ESM2M": {"name": "GFDL-ESM2M"},
+            },
             "metric": {"carbon": {"name": "carbon"}},
         },
         "RESULTS": {
-            "E3SM": {"carbon": {"overall score": 0.11, "bias": 0.56}},
-            "CESM": {"carbon": {"overall score": 0.05, "bias": 0.72}},
+            "E3SM": {
+                "carbon": {
+                    "overall score": 0.11,
+                    "bias": 0.56,
+                    "attributes": {
+                        "score": "ILAMB scoring system",
+                    },
+                },
+            },
+            "CESM": {
+                "carbon": {"overall score": 0.05, "bias": 0.72},
+            },
+            "GFDL-ESM2M": {
+                "carbon": {"overall score": 0.35, "bias": 0.37},
+                "attributes": {
+                    "score": "ILAMB scoring system",
+                },
+            },
+            "attributes": {
+                "package": "ilamb",
+                "overall score": {},
+            },
         },
     }
 
@@ -229,7 +261,13 @@ def test_metric_merge():
         },
         "RESULTS": {
             "E3SM": {
-                "carbon": {"overall score": 0.11, "bias": 0.56},
+                "carbon": {
+                    "overall score": 0.11,
+                    "bias": 0.56,
+                    "attributes": {
+                        "score": "ILAMB scoring system",
+                    },
+                },
                 "NinoSstDiversity_2": {},
                 "BiasTauxLonRmse": {},
             },
@@ -241,7 +279,17 @@ def test_metric_merge():
             "GFDL-ESM2M": {
                 "NinoSstDiversity_2": -75,
                 "BiasTauxLonRmse": 5.959564210237568,
-                "carbon": {},
+                "attributes": {
+                    "NinoSstDiversity_2": "Nino SST diversity",
+                    "BiasTauxLonRmse": "Bias Taux RMSE",
+                    "score": "ILAMB scoring system",
+                },
+                "carbon": {"overall score": 0.35, "bias": 0.37},
+            },
+            "attributes": {
+                "package": "ilamb",
+                "NinoSstDiversity_2": {},
+                "overall score": {},
             },
         },
         "PROVENANCE": None,

packages/ref-metrics-pmp/src/cmip_ref_metrics_pmp/pmp_driver.py

@@ -87,7 +87,7 @@ def process_json_result(
 
     # Remove the "attributes" key from the RESULTS
     # This isn't standard CMEC output, but it is what PMP produces
-    results = _remove_nested_key(json_result["RESULTS"], "attributes")
+    results = json_result["RESULTS"]
 
     cmec_metric["RESULTS"] = results
     cmec_metric["DIMENSIONS"] = dimensions