Iterate on tesrts

src/napari_easy_augment_batch_dl/_tests/test_deep_learning_project.py

@@ -1,20 +1,47 @@
 
 from napari_easy_augment_batch_dl.deep_learning_project import DeepLearningProject, ImageType
-from enum import Enum
 import os
+import shutil 
+import pytest
+from pathlib import Path
+import json
+from napari_easy_augment_batch_dl.deep_learning_project import x1y1x2y2_to_tltrblbr
 
-
-def test_deep_learning_project():
+@pytest.fixture(scope="module", autouse=True)
+def setup_and_cleanup():
     print('cwd is',os.getcwd())
 
     test_data_path = r'./test_data'
-    parent_path = os.path.join(test_data_path, ImageType.GRAY_SCALE_2D.value)
 
+    parent_path_gray_scale_2d = os.path.join(test_data_path, ImageType.GRAY_SCALE_2D.value)
+    parent_path_gray_scale_2d_temp = os.path.join(test_data_path, 'temp', ImageType.GRAY_SCALE_2D.value)
+    # Create the temporary directory
+    os.makedirs(parent_path_gray_scale_2d_temp, exist_ok=True)
     # confirm parent path exists
-    assert os.path.exists(parent_path), f"Parent path {parent_path} does not exist"
-
-    deep_learning_project = DeepLearningProject(parent_path, 2)
+    assert os.path.exists(parent_path_gray_scale_2d), f"Parent path {parent_path_gray_scale_2d} does not exist"
+    # Copy the parent_path to the temporary directory
+    shutil.copytree(parent_path_gray_scale_2d, parent_path_gray_scale_2d_temp, dirs_exist_ok=True)
+    # Run the test on the temporary directory
+    deep_learning_project_gray_scale_2d = DeepLearningProject(parent_path_gray_scale_2d_temp, 2)
 
+    parent_path_rgb_2d = os.path.join(test_data_path, ImageType.RGB_2D.value)
+    parent_path_rgb_2d_temp = os.path.join(test_data_path, 'temp', ImageType.RGB_2D.value)
+    # Create the temporary directory
+    os.makedirs(parent_path_rgb_2d_temp, exist_ok=True)
+    # confirm parent path exists
+    assert os.path.exists(parent_path_rgb_2d), f"Parent path {parent_path_rgb_2d} does not exist"
+    # Copy the parent_path to the temporary directory
+    shutil.copytree(parent_path_rgb_2d, parent_path_rgb_2d_temp, dirs_exist_ok=True)
+    deep_learning_project_rgb_2d = DeepLearningProject(parent_path_rgb_2d_temp, 2)
+
+    yield deep_learning_project_gray_scale_2d, deep_learning_project_rgb_2d
+
+    # Cleanup after test
+    shutil.rmtree(parent_path_gray_scale_2d_temp, ignore_errors=True)
+    shutil.rmtree(parent_path_rgb_2d_temp, ignore_errors=True)
+
+def test_deep_learning_project(setup_and_cleanup):
+    deep_learning_project, _ = setup_and_cleanup
     print('deep_learning_project', type(deep_learning_project))
     print('number classes is', deep_learning_project.num_classes)
 
@@ -24,7 +51,87 @@ def test_deep_learning_project():
 
     print('number of images is', len(deep_learning_project.image_file_list))
     print('image names are', deep_learning_project.image_file_list)
-    
+
     for i in range(len(deep_learning_project.image_list)):
         image = deep_learning_project.image_list[i]
-        print('image', i, type(image), image.shape)
+        print('image', i, type(image), image.shape)
+
+def test_augmentation_project(setup_and_cleanup):
+    perform_augmentation(setup_and_cleanup[0])
+    perform_augmentation(setup_and_cleanup[1], do_color_jitter=True)
+
+def perform_augmentation(deep_learning_project, do_horizontal_flip=True, do_vertical_flip=True, do_random_rotate90=True, do_random_sized_crop=True, 
+        do_random_brightness_contrast=True, do_random_gamma=False, do_color_jitter=False, do_elastic_transform=False):
+
+    json_names = list(Path(deep_learning_project.image_label_paths[0]).glob('*.json'))
+    image_name = deep_learning_project.image_file_list[1]
+    image_base_name = image_name.name.split('.')[0]
+    #image_base_name = Path(image_name).stem
+
+    json_names_ = [x for x in json_names if image_base_name in x.name]
+
+    print('image_base_name', image_base_name)
+    print('json_names', json_names)
+    print('filtered json_names', json_names_)
+
+    json_name = json_names_[0]
+
+    boxes = []
+    with open(json_name, 'r') as f:
+        json_ = json.load(f)
+        #print(json_)
+
+        x1= json_['bbox'][0]
+        y1= json_['bbox'][1]
+        x2= json_['bbox'][2]
+        y2= json_['bbox'][3]
+
+        bbox = x1y1x2y2_to_tltrblbr(x1, y1, x2, y2, 1)
+        boxes.append(bbox)        
+
+
+    deep_learning_project.perform_augmentation(boxes, num_patches = 100, patch_size=256, do_horizontal_flip=do_horizontal_flip, do_vertical_flip=do_vertical_flip,
+                                               do_random_rotate90=do_random_rotate90, do_random_sized_crop=do_random_sized_crop, do_random_brightness_contrast=do_random_brightness_contrast,
+                                               do_random_gamma=do_random_gamma, do_color_jitter=do_color_jitter, do_elastic_transform=do_elastic_transform)
+
+
+
+'''
+def test_deep_learning_project():
+    print('cwd is',os.getcwd())
+
+    test_data_path = r'./test_data'
+    parent_path = os.path.join(test_data_path, ImageType.GRAY_SCALE_2D.value)
+    parent_path_temp = os.path.join(test_data_path, 'temp', ImageType.GRAY_SCALE_2D.value)
+
+    # Create the temporary directory
+    os.makedirs(parent_path_temp, exist_ok=True)
+
+    # confirm parent path exists
+    assert os.path.exists(parent_path), f"Parent path {parent_path} does not exist"
+
+    try:
+        # Copy the parent_path to the temporary directory
+        shutil.copytree(parent_path, parent_path_temp, dirs_exist_ok=True)
+
+        # Run the test on the temporary directory
+        deep_learning_project = DeepLearningProject(parent_path_temp, 2)
+
+        print('deep_learning_project', type(deep_learning_project))
+        print('number classes is', deep_learning_project.num_classes)
+
+        num_images = len(deep_learning_project.image_file_list)
+
+        assert num_images == 2, f"Expected 2 images, found {num_images}"
+
+        print('number of images is', len(deep_learning_project.image_file_list))
+        print('image names are', deep_learning_project.image_file_list)
+
+        for i in range(len(deep_learning_project.image_list)):
+            image = deep_learning_project.image_list[i]
+            print('image', i, type(image), image.shape)
+    finally:
+        # Clean up the temporary directory
+        #shutil.rmtree(os.path.join(test_data_path, 'temp'))
+        pass
+'''