fixed training for annotated examples where fields are annotated but …

…form type is not
TeamHG-Memex · Feb 1, 2016 · 5b5086c · 5b5086c
1 parent 167878d
commit 5b5086c
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 1 deletion.
diff --git a/formasaurus/annotation.py b/formasaurus/annotation.py
@@ -42,6 +42,10 @@ def fields_annotated(self):
             for v in self.fields.values()
         )
 
+    @property
+    def form_annotated(self):
+        return self.type != self.form_schema.na_value
+
     @property
     def fields_partially_annotated(self):
         """

diff --git a/formasaurus/fieldtype_model.py b/formasaurus/fieldtype_model.py
@@ -63,7 +63,10 @@ def log(msg):
         if verbose:
             print(msg)
 
-    annotations = [a for a in annotations if a.fields_annotated]
+    annotations = [
+        a for a in annotations
+        if a.fields_annotated and (a.form_annotated or not use_precise_form_types)
+    ]
     log("Training on {} forms".format(len(annotations)))
 
     if use_precise_form_types: