Merge pull request #49 from SFI-Visual-Intelligence/christian/test-model-metric-data

Create tests for load_model/metric/data, closes #46
LGTM

Author: Seilmast

.gitignore

@@ -7,6 +7,7 @@ _build/
 bin/*
 wandb/*
 wandb_api.py
+doc/autoapi
 
 #Magnus specific
 job*

tests/test_dataloaders.py

@@ -1,8 +1,41 @@
-from utils.dataloaders.usps_0_6 import USPSDataset0_6
+from pathlib import Path
+
+import numpy as np
+import pytest
+import torch
+from PIL import Image
+from torchvision import transforms
+
+from utils.dataloaders import MNISTDataset0_3, USPSDataset0_6, USPSH5_Digit_7_9_Dataset
+from utils.load_data import load_data
+
+
+@pytest.mark.parametrize(
+    "data_name, expected",
+    [
+        ("usps_0-6", USPSDataset0_6),
+        ("usps_7-9", USPSH5_Digit_7_9_Dataset),
+        ("mnist_0-3", MNISTDataset0_3),
+        # TODO: Add more datasets here
+    ],
+)
+def test_load_data(data_name, expected):
+    dataset = load_data(
+        data_name,
+        data_path=Path("data"),
+        download=True,
+        transform=transforms.ToTensor(),
+    )
+    assert isinstance(dataset, expected)
+    assert len(dataset) > 0
+    assert isinstance(dataset[0], tuple)
+    assert isinstance(dataset[0][0], torch.Tensor)
+    assert isinstance(
+        dataset[0][1], (int, torch.Tensor, np.ndarray)
+    )  # Should probably restrict this to only int or one-hot encoded tensor or array for consistency.
 
 
 def test_uspsdataset0_6():
-    from pathlib import Path
     from tempfile import TemporaryDirectory
 
     import h5py

utils/dataloaders/uspsh5_7_9.py

@@ -50,6 +50,8 @@ def __init__(self, data_path, train=False, transform=None):
         self.filepath = path / self.filename
         self.transform = transform
         self.mode = "train" if train else "test"
+        self.h5_path = data_path / self.filename
+
         # Load the dataset from the HDF5 file
         with h5py.File(self.filepath, "r") as hf:
             images = hf[self.mode]["data"][:]