[Bugfix] Temporarily modify single table demo data link (#121)

* change demo data link * Fix col names --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
hitsz-ids · Jan 20, 2024 · 1bbf3f5 · 1bbf3f5
1 parent 9ac9988
commit 1bbf3f5
Show file tree

Hide file tree

Showing 4 changed files with 15 additions and 13 deletions.
diff --git a/sdgx/utils.py b/sdgx/utils.py
@@ -65,8 +65,10 @@ def download_demo_data(data_dir: str | Path = "./dataset") -> Path:
         # Download from datahub
         demo_data_path.parent.mkdir(parents=True, exist_ok=True)
 
-        logger.info("Downloading demo data from datahub.io to {}".format(demo_data_path))
-        url = "https://datahub.io/machine-learning/adult/r/adult.csv"
+        logger.info("Downloading demo data from github data source to {}".format(demo_data_path))
+        url = (
+            "https://raw.githubusercontent.com/saravrajavelu/Adult-Income-Analysis/master/adult.csv"
+        )
         urllib.request.urlretrieve(url, demo_data_path)
 
     return demo_data_path
@@ -93,9 +95,9 @@ def get_demo_single_table(data_dir: str | Path = "./dataset"):
         "occupation",
         "relationship",
         "race",
-        "sex",
+        "gender",
         "native-country",
-        "class",
+        "income",
     ]
     return pd_obj, discrete_cols
 

diff --git a/tests/data_models/inspector/test_discrete.py b/tests/data_models/inspector/test_discrete.py
@@ -26,9 +26,9 @@ def test_inspector(inspector: DiscreteInspector, raw_data):
             "occupation",
             "relationship",
             "race",
-            "sex",
+            "gender",
             "native-country",
-            "class",
+            "income",
         ]
     )
 

diff --git a/tests/data_models/inspector/test_numeric.py b/tests/data_models/inspector/test_numeric.py
@@ -19,7 +19,7 @@ def test_inspector(inspector: NumericInspector, raw_data):
     assert inspector.ready
     assert inspector.numeric_columns
     assert sorted(inspector.inspect()["numeric_columns"]) == sorted(
-        ["education-num", "fnlwgt", "hoursperweek", "age", "capitalgain", "capitalloss"]
+        ["educational-num", "fnlwgt", "hours-per-week", "age", "capital-gain", "capital-loss"]
     )
 
 

diff --git a/tests/dataloader/test_dataloader.py b/tests/dataloader/test_dataloader.py
@@ -24,17 +24,17 @@ def test_demo_dataloader(dataloader_builder: DataLoader, cacher, demo_single_tab
                 "workclass",
                 "fnlwgt",
                 "education",
-                "education-num",
+                "educational-num",
                 "marital-status",
                 "occupation",
                 "relationship",
                 "race",
-                "sex",
-                "capitalgain",
-                "capitalloss",
-                "hoursperweek",
+                "gender",
+                "capital-gain",
+                "capital-loss",
+                "hours-per-week",
                 "native-country",
-                "class",
+                "income",
             ]
         )
     )