closurizer 0.1.7__py3-none-any.whl → 0.2.1__py3-none-any.whl

closurizer/closurizer.py

@@ -15,11 +15,24 @@ def _cut_left_join(ltable, rtable, field, attribute):
                         lkey=field,
                         rkey="id")
 
+def _length(value):
+    if value is None:
+        return 0
+    else:
+        return len(value.split("|"))
+
+def _length_of_field_values(rec, fields):
+    value = 0
+    for field in fields:
+        if (field_value := rec[field]) is not None:
+            value += _length(field_value)
+    return value
 
 def add_closure(kg_archive: str,
                 closure_file: str,
                 output_file: str,
-                fields: List[str] = None
+                fields: List[str] = ['subject', 'object'],
+                evidence_fields: List[str] = ['has_evidence', 'publications', 'primary_knowledge_source']
                 ):
     print("Generating closure KG...")
     print(f"kg_archive: {kg_archive}")
@@ -28,6 +41,9 @@ def add_closure(kg_archive: str,
     if fields is None or len(fields) == 0:
         fields = ['subject', 'object']
 
+    if evidence_fields is None or len(evidence_fields) == 0:
+        evidence_fields = ['has_evidence', 'publications', 'primary_knowledge_source', 'provided_by']
+
     print(f"fields: {','.join(fields)}")
     print(f"output_file: {output_file}")
 
@@ -79,6 +95,11 @@ def add_closure(kg_archive: str,
         edges = etl.leftjoin(edges, (etl.cut(nodes, ["id", "name"]).rename("name", f"{field}_label")), lkey=field,
                              rkey="id")
 
+    print("Adding evidence counts...")
+
+    edges = etl.addfield(edges, 'evidence_count',
+                         lambda rec: _length_of_field_values(rec, evidence_fields))
+
     print("Denormalizing...")
     etl.totsv(edges, f"{output_file}")
 

{closurizer-0.1.7.dist-info → closurizer-0.2.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: closurizer
-Version: 0.1.7
+Version: 0.2.1
 Summary: Add closure expansion fields to kgx files following the Golr pattern
 Author: Kevin Schaper
 Author-email: kevin@tislab.org

closurizer-0.2.1.dist-info/RECORD

@@ -0,0 +1,6 @@
+closurizer/cli.py,sha256=2SsVXscVJFgwelfZvHKQQQtTU-9pGaA2RJnK2BaNVc8,641
+closurizer/closurizer.py,sha256=UAZJ2XCwSqXhp5AgBfUu6s_H72CtZK2H3Jc-746cVrw,4352
+closurizer-0.2.1.dist-info/METADATA,sha256=sRqV2_sM2-iB_hzHafpjnm8X7IXuN2oYOLgjDR1AWfg,575
+closurizer-0.2.1.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+closurizer-0.2.1.dist-info/entry_points.txt,sha256=MnAVu1lgP6DqDb3BZGNzVs2AnDMsp4sThi3ccWbONFo,50
+closurizer-0.2.1.dist-info/RECORD,,

{closurizer-0.1.7.dist-info → closurizer-0.2.1.dist-info}/WHEEL

@@ -1,4 +1,4 @@
 Wheel-Version: 1.0
-Generator: poetry-core 1.3.2
+Generator: poetry-core 1.6.1
 Root-Is-Purelib: true
 Tag: py3-none-any

closurizer-0.1.7.dist-info/RECORD

@@ -1,6 +0,0 @@
-closurizer/cli.py,sha256=2SsVXscVJFgwelfZvHKQQQtTU-9pGaA2RJnK2BaNVc8,641
-closurizer/closurizer.py,sha256=rmrOJFh83eGR-hfVMytDNc6zd3np30qSIobmHOZYsEk,3590
-closurizer-0.1.7.dist-info/entry_points.txt,sha256=MnAVu1lgP6DqDb3BZGNzVs2AnDMsp4sThi3ccWbONFo,50
-closurizer-0.1.7.dist-info/WHEEL,sha256=vxFmldFsRN_Hx10GDvsdv1wroKq8r5Lzvjp6GZ4OO8c,88
-closurizer-0.1.7.dist-info/METADATA,sha256=9yYJxRgs8D9NVUQtgW94NEIiaMFj1qiE4ci0T3e4v80,575
-closurizer-0.1.7.dist-info/RECORD,,