Intro dummy table for speedup models case

tests/conftest.py

@@ -2,9 +2,12 @@
 
 os.environ["SDG_NDARRAY_CACHE_ROOT"] = "/tmp/sdgx/ndarray_cache"
 
+import random
 import shutil
+import string
 from functools import partial
 
+import pandas as pd
 import pytest
 
 from sdgx.data_connectors.csv_connector import CsvConnector
@@ -18,6 +21,49 @@
 DATA_DIR = os.path.join(_HERE, "dataset")
 
 
+def ramdon_str():
+    return "".join(random.choice(string.ascii_letters) for _ in range(10))
+
+
+@pytest.fixture
+def dummy_single_table_path(tmp_path):
+    dummy_size = 10
+    role_set = ["admin", "user", "guest"]
+
+    df = pd.DataFrame(
+        {
+            "role": [random.choice(role_set) for _ in range(dummy_size)],
+            "name": [ramdon_str() for _ in range(dummy_size)],
+            "feature_x": [random.random() for _ in range(dummy_size)],
+            "feature_y": [random.random() for _ in range(dummy_size)],
+            "feature_z": [random.random() for _ in range(dummy_size)],
+        }
+    )
+    save_path = tmp_path / "dummy.csv"
+    df.to_csv(save_path, index=False, header=True)
+    yield save_path
+    save_path.unlink()
+
+
+@pytest.fixture
+def dummy_single_table_data_connector(dummy_single_table_path):
+    yield CsvConnector(
+        path=dummy_single_table_path,
+    )
+
+
+@pytest.fixture
+def dummy_single_table_data_loader(dummy_single_table_data_connector, cacher_kwargs):
+    d = DataLoader(dummy_single_table_data_connector, cacher_kwargs=cacher_kwargs)
+    yield d
+    d.finalize()
+
+
+@pytest.fixture
+def dummy_single_table_metadata(dummy_single_table_data_loader):
+    yield Metadata.from_dataloader(dummy_single_table_data_loader)
+
+
 @pytest.fixture
 def demo_single_table_path():
     yield download_demo_data(DATA_DIR).as_posix()

tests/models/test_copula.py

@@ -1,13 +1,26 @@
 from pathlib import Path
 
+import pandas as pd
+import pytest
+
 from sdgx.models.statistics.single_table.copula import GaussianCopulaSynthesizer
 from sdgx.utils import get_demo_single_table
 
 
-def test_gaussian_copula(demo_single_table_path):
-    demo_data, discrete_cols = get_demo_single_table(Path(demo_single_table_path).parent)
+@pytest.fixture
+def dummy_data(dummy_single_table_path):
+    yield pd.read_csv(dummy_single_table_path)
+
+
+@pytest.fixture
+def discrete_cols(dummy_data):
+    yield [col for col in dummy_data.columns if not col.startswith("feature")]
+
+
+def test_gaussian_copula(dummy_data, discrete_cols):
     model = GaussianCopulaSynthesizer(discrete_cols)
-    model.fit(demo_data)
+    model.fit(dummy_data)
 
     sampled_data = model.sample(10)
     assert len(sampled_data) == 10
+    assert sampled_data.columns.tolist() == dummy_data.columns.tolist()

tests/models/test_ctgan.py

@@ -17,17 +17,27 @@ def save_model_dir(tmp_path):
     shutil.rmtree(dirname, ignore_errors=True)
 
 
+def assert_sampled_data(dummy_single_table_data_loader, sampled_data, count):
+    assert len(sampled_data) == count
+    assert sampled_data.columns.tolist() == dummy_single_table_data_loader.columns()
+
+
 def test_ctgan(
     ctgan: CTGANSynthesizerModel,
-    demo_single_table_metadata,
-    demo_single_table_data_loader,
+    dummy_single_table_metadata,
+    dummy_single_table_data_loader,
     save_model_dir,
 ):
-    ctgan.fit(demo_single_table_metadata, demo_single_table_data_loader)
-    ctgan.sample(10)
+    ctgan.fit(dummy_single_table_metadata, dummy_single_table_data_loader)
+    sampled_data = ctgan.sample(10)
+    assert_sampled_data(dummy_single_table_data_loader, sampled_data, 10)
+
     ctgan.save(save_model_dir)
+    assert save_model_dir.exists()
+
     model = CTGANSynthesizerModel.load(save_model_dir)
-    model.sample(10)
+    sampled_data = model.sample(10)
+    assert_sampled_data(dummy_single_table_data_loader, sampled_data, 10)
 
 
 if __name__ == "__main__":