[hotfix] 'No numeric types to aggregate'

superset/models/core.py

@@ -575,13 +575,7 @@ class Database(Model, AuditMixinNullable):
     def get_df(self, sql, schema):
         sql = sql.strip().strip(';')
         eng = self.get_sqla_engine(schema=schema)
-
-        conn = eng.raw_connection()
-        cur = conn.cursor()
-        cur.execute(sql, **self.db_engine_spec.cursor_execute_kwargs)
-
-        cols = [col[0] for col in cur.description]
-        df = pd.DataFrame(list(cur.fetchall()), columns=cols)
+        df = pd.read_sql(sql, eng)
 
         def needs_conversion(df_series):
             if df_series.empty: