Update id_column and id_column_type names in PyTest (#347)

* update id_column names in pytest Signed-off-by: Sarah Yurick <[email protected]> * edit Signed-off-by: Sarah Yurick <[email protected]> * edit config Signed-off-by: Sarah Yurick <[email protected]> --------- Signed-off-by: Sarah Yurick <[email protected]>
NVIDIA · Nov 7, 2024 · 1e3ccc4 · 1e3ccc4
1 parent 01bda47
commit 1e3ccc4
Showing 1 changed file with 6 additions and 4 deletions.
diff --git a/tests/test_semdedup.py b/tests/test_semdedup.py
@@ -64,15 +64,17 @@ def test_sem_dedup(
         cache_dir = os.path.join(tmpdir, "test_sem_dedup_cache")
         config = SemDedupConfig(
             cache_dir=cache_dir,
-            id_col_name="id",
-            id_col_type="int",
-            input_column="text",
             seed=42,
             n_clusters=3,
             eps_thresholds=[0.10],
             eps_to_extract=0.10,
         )
-        sem_duplicates = SemDedup(config=config)
+        sem_duplicates = SemDedup(
+            config=config,
+            input_column="text",
+            id_column="id",
+            id_column_type="int",
+        )
         result = sem_duplicates(dedup_data)
         result_df = result.df.compute()
         duplicate_docs = [2, 3, 4, 200, 300]