remove time leakage columns

snap-stanford · Jan 29, 2024 · a3f07d0 · a3f07d0
1 parent d5fb98a
commit a3f07d0
Showing 1 changed file with 4 additions and 1 deletion.
diff --git a/relbench/datasets/stackex.py b/relbench/datasets/stackex.py
@@ -47,6 +47,7 @@ def make_db(self) -> Database:
             columns=["Reputation", "Views", "UpVotes", "DownVotes", "LastAccessDate"],
             inplace=True,
         )
+
         posts.drop(
             columns=[
                 "ViewCount",
@@ -58,6 +59,9 @@ def make_db(self) -> Database:
                 "LastEditDate",
                 "LastActivityDate",
                 "Score",
+                "LastEditorDisplayName",
+                "LastEditorUserId",
+                "AcceptedAnswerId"
             ],
             inplace=True,
         )
@@ -131,7 +135,6 @@ def make_db(self) -> Database:
             df=pd.DataFrame(posts),
             fkey_col_to_pkey_table={
                 "OwnerUserId": "users",
-                "LastEditorUserId": "users",
                 "ParentId": "posts",  # notice the self-reference
             },
             pkey_col="Id",