File size: 15,821 Bytes
18a82fb | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 | #!/usr/bin/env python3
"""
Comprehensive test script for transforms.py
Tests all functions and classes independently
Run from project root: python -c "from src.data.transforms import *; exec(open('test_transforms_main.py').read())"
"""
import torch
import numpy as np
from PIL import Image
import matplotlib.pyplot as plt
from typing import Dict, List
import traceback
import time
# Create test images of different sizes and types
def create_test_images() -> Dict[str, Image.Image]:
"""Create various test images"""
test_images = {}
# Standard test image
test_images['standard'] = Image.fromarray(
np.random.randint(0, 255, (300, 300, 3), dtype=np.uint8)
)
# Small image
test_images['small'] = Image.fromarray(
np.random.randint(0, 255, (50, 50, 3), dtype=np.uint8)
)
# Large image
test_images['large'] = Image.fromarray(
np.random.randint(0, 255, (800, 600, 3), dtype=np.uint8)
)
# Non-square image
test_images['rectangle'] = Image.fromarray(
np.random.randint(0, 255, (400, 200, 3), dtype=np.uint8)
)
# Grayscale converted to RGB
gray_array = np.random.randint(0, 255, (200, 200), dtype=np.uint8)
test_images['grayscale'] = Image.fromarray(gray_array, 'L').convert('RGB')
return test_images
def test_basic_transforms():
"""Test basic transform functions"""
print("=" * 60)
print("TESTING BASIC TRANSFORMS")
print("=" * 60)
test_images = create_test_images()
results = {}
# Test different input sizes
input_sizes = [224, 256, 384]
for size in input_sizes:
print(f"\n--- Testing input size: {size} ---")
# Test get_train_transforms
try:
for level in ['light', 'medium', 'heavy']:
transform = get_train_transforms(input_size=size, augmentation_level=level)
for img_name, img in test_images.items():
tensor = transform(img)
expected_shape = (3, size, size)
assert tensor.shape == expected_shape, f"Wrong shape: {tensor.shape} vs {expected_shape}"
assert tensor.dtype == torch.float32, f"Wrong dtype: {tensor.dtype}"
print(f"β get_train_transforms({level}) works for all test images")
results[f'train_transforms_{size}'] = "PASS"
except Exception as e:
print(f"β get_train_transforms failed: {e}")
results[f'train_transforms_{size}'] = f"FAIL: {e}"
# Test get_val_transforms
try:
val_transform = get_val_transforms(input_size=size)
for img_name, img in test_images.items():
tensor = val_transform(img)
expected_shape = (3, size, size)
assert tensor.shape == expected_shape, f"Wrong shape: {tensor.shape} vs {expected_shape}"
assert tensor.dtype == torch.float32, f"Wrong dtype: {tensor.dtype}"
print(f"β get_val_transforms works for all test images")
results[f'val_transforms_{size}'] = "PASS"
except Exception as e:
print(f"β get_val_transforms failed: {e}")
results[f'val_transforms_{size}'] = f"FAIL: {e}"
# Test get_inference_transforms
try:
inference_transform = get_inference_transforms(224)
tensor = inference_transform(test_images['standard'])
assert tensor.shape == (3, 224, 224), f"Wrong inference shape: {tensor.shape}"
print("β get_inference_transforms works")
results['inference_transforms'] = "PASS"
except Exception as e:
print(f"β get_inference_transforms failed: {e}")
results['inference_transforms'] = f"FAIL: {e}"
return results
def test_model_specific_transforms():
"""Test model-specific transform function"""
print("\n" + "=" * 60)
print("TESTING MODEL-SPECIFIC TRANSFORMS")
print("=" * 60)
results = {}
# Test different model configurations
test_models = [
('resnet50', 224),
('efficientnet-b2', 260), # Assuming this is configured
('convnext-tiny-384', 384), # Assuming this is configured
('unknown_model', 224) # Should fallback to default
]
for model_name, expected_size in test_models:
print(f"\n--- Testing model: {model_name} ---")
try:
# Test training transforms
train_transform = get_transforms_for_model(model_name, is_training=True)
val_transform = get_transforms_for_model(model_name, is_training=False)
test_img = Image.fromarray(np.random.randint(0, 255, (300, 300, 3), dtype=np.uint8))
train_tensor = train_transform(test_img)
val_tensor = val_transform(test_img)
print(f"β Train transform output: {train_tensor.shape}")
print(f"β Val transform output: {val_tensor.shape}")
# Note: We don't assert exact size since config might use defaults
assert train_tensor.shape[0] == 3, "Wrong number of channels"
assert val_tensor.shape[0] == 3, "Wrong number of channels"
assert train_tensor.shape[1] == train_tensor.shape[2], "Not square"
assert val_tensor.shape[1] == val_tensor.shape[2], "Not square"
results[f'model_{model_name}'] = "PASS"
except Exception as e:
print(f"β Model {model_name} failed: {e}")
results[f'model_{model_name}'] = f"FAIL: {e}"
return results
def test_advanced_augmentations():
"""Test MixUp, CutMix, and RandAugment"""
print("\n" + "=" * 60)
print("TESTING ADVANCED AUGMENTATIONS")
print("=" * 60)
results = {}
# Test MixUp
print("\n--- Testing MixUp ---")
try:
mixup = MixUpTransform(alpha=1.0, num_classes=5)
batch_size = 4
images = torch.randn(batch_size, 3, 224, 224)
labels = torch.tensor([0, 1, 2, 3])
mixed_images, labels_a, labels_b, lam = mixup(images, labels)
assert mixed_images.shape == images.shape, f"MixUp shape mismatch: {mixed_images.shape}"
assert 0 <= lam <= 1, f"Lambda out of range: {lam}"
assert labels_a.shape == labels.shape, "Labels_a shape mismatch"
assert labels_b.shape == labels.shape, "Labels_b shape mismatch"
print(f"β MixUp works: lambda={lam:.3f}, output_shape={mixed_images.shape}")
results['mixup'] = "PASS"
except Exception as e:
print(f"β MixUp failed: {e}")
results['mixup'] = f"FAIL: {e}"
# Test CutMix
print("\n--- Testing CutMix ---")
try:
cutmix = CutMixTransform(alpha=1.0, num_classes=5)
batch_size = 4
images = torch.randn(batch_size, 3, 224, 224)
labels = torch.tensor([0, 1, 2, 3])
mixed_images, labels_a, labels_b, lam = cutmix(images, labels)
assert mixed_images.shape == images.shape, f"CutMix shape mismatch: {mixed_images.shape}"
assert 0 <= lam <= 1, f"Lambda out of range: {lam}"
assert labels_a.shape == labels.shape, "Labels_a shape mismatch"
assert labels_b.shape == labels.shape, "Labels_b shape mismatch"
print(f"β CutMix works: lambda={lam:.3f}, output_shape={mixed_images.shape}")
results['cutmix'] = "PASS"
except Exception as e:
print(f"β CutMix failed: {e}")
results['cutmix'] = f"FAIL: {e}"
# Test RandAugment
print("\n--- Testing RandAugment ---")
try:
randaugment = RandAugmentTransform(n=2, m=10)
test_img = Image.fromarray(np.random.randint(0, 255, (224, 224, 3), dtype=np.uint8))
augmented_img = randaugment(test_img)
assert isinstance(augmented_img, Image.Image), "RandAugment should return PIL Image"
assert augmented_img.size == test_img.size, "RandAugment changed image size"
print(f"β RandAugment works: input_size={test_img.size}, output_size={augmented_img.size}")
results['randaugment'] = "PASS"
except Exception as e:
print(f"β RandAugment failed: {e}")
results['randaugment'] = f"FAIL: {e}"
return results
def test_advanced_train_transforms():
"""Test advanced training transforms with RandAugment"""
print("\n" + "=" * 60)
print("TESTING ADVANCED TRAIN TRANSFORMS")
print("=" * 60)
results = {}
try:
# Test without RandAugment
transform_normal = get_advanced_train_transforms(
input_size=224,
use_randaugment=False
)
# Test with RandAugment
transform_rand = get_advanced_train_transforms(
input_size=224,
use_randaugment=True,
randaugment_n=2,
randaugment_m=5
)
test_img = Image.fromarray(np.random.randint(0, 255, (300, 300, 3), dtype=np.uint8))
tensor_normal = transform_normal(test_img)
tensor_rand = transform_rand(test_img)
assert tensor_normal.shape == (3, 224, 224), f"Normal transform wrong shape: {tensor_normal.shape}"
assert tensor_rand.shape == (3, 224, 224), f"RandAugment transform wrong shape: {tensor_rand.shape}"
print(f"β Advanced transforms work: normal={tensor_normal.shape}, randaugment={tensor_rand.shape}")
results['advanced_transforms'] = "PASS"
except Exception as e:
print(f"β Advanced transforms failed: {e}")
results['advanced_transforms'] = f"FAIL: {e}"
return results
def test_denormalize():
"""Test denormalization utility"""
print("\n" + "=" * 60)
print("TESTING DENORMALIZATION")
print("=" * 60)
results = {}
try:
# Create normalized tensor
transform = get_val_transforms(224)
test_img = Image.fromarray(np.random.randint(0, 255, (224, 224, 3), dtype=np.uint8))
normalized_tensor = transform(test_img)
# Denormalize
denorm = DeNormalize()
denormalized_tensor = denorm(normalized_tensor)
assert denormalized_tensor.shape == normalized_tensor.shape, "Denorm changed shape"
assert denormalized_tensor.dtype == torch.float32, "Denorm changed dtype"
# Check if values are in reasonable range (0-1 for images)
min_val = denormalized_tensor.min().item()
max_val = denormalized_tensor.max().item()
print(f"β Denormalization works: shape={denormalized_tensor.shape}")
print(f" Normalized range: [{normalized_tensor.min():.3f}, {normalized_tensor.max():.3f}]")
print(f" Denormalized range: [{min_val:.3f}, {max_val:.3f}]")
results['denormalize'] = "PASS"
except Exception as e:
print(f"β Denormalization failed: {e}")
results['denormalize'] = f"FAIL: {e}"
return results
def test_performance():
"""Test transform performance"""
print("\n" + "=" * 60)
print("TESTING PERFORMANCE")
print("=" * 60)
results = {}
try:
transform = get_train_transforms(224, 'medium')
test_img = Image.fromarray(np.random.randint(0, 255, (224, 224, 3), dtype=np.uint8))
# Time multiple transforms
n_iterations = 100
start_time = time.time()
for _ in range(n_iterations):
_ = transform(test_img)
end_time = time.time()
avg_time = (end_time - start_time) / n_iterations * 1000 # ms
print(f"β Performance test: {avg_time:.2f}ms per transform (avg of {n_iterations} iterations)")
if avg_time < 50: # Less than 50ms is good
results['performance'] = "PASS"
else:
results['performance'] = f"SLOW: {avg_time:.2f}ms"
except Exception as e:
print(f"β Performance test failed: {e}")
results['performance'] = f"FAIL: {e}"
return results
def test_edge_cases():
"""Test edge cases and error handling"""
print("\n" + "=" * 60)
print("TESTING EDGE CASES")
print("=" * 60)
results = {}
# Test very small image
try:
small_img = Image.fromarray(np.random.randint(0, 255, (10, 10, 3), dtype=np.uint8))
transform = get_val_transforms(224)
tensor = transform(small_img)
assert tensor.shape == (3, 224, 224), "Small image transform failed"
print("β Very small image (10x10) handled correctly")
results['small_image'] = "PASS"
except Exception as e:
print(f"β Small image test failed: {e}")
results['small_image'] = f"FAIL: {e}"
# Test extreme aspect ratio
try:
wide_img = Image.fromarray(np.random.randint(0, 255, (50, 500, 3), dtype=np.uint8))
transform = get_val_transforms(224)
tensor = transform(wide_img)
assert tensor.shape == (3, 224, 224), "Wide image transform failed"
print("β Extreme aspect ratio (1:10) handled correctly")
results['aspect_ratio'] = "PASS"
except Exception as e:
print(f"β Aspect ratio test failed: {e}")
results['aspect_ratio'] = f"FAIL: {e}"
# Test invalid augmentation level
try:
transform = get_train_transforms(224, 'invalid_level')
# Should fallback to medium
print("β Invalid augmentation level handled (fallback to medium)")
results['invalid_aug_level'] = "PASS"
except Exception as e:
print(f"β Invalid augmentation level test failed: {e}")
results['invalid_aug_level'] = f"FAIL: {e}"
return results
def print_final_summary(all_results: Dict[str, Dict]):
"""Print comprehensive test summary"""
print("\n" + "=" * 60)
print("FINAL TEST SUMMARY")
print("=" * 60)
total_tests = 0
passed_tests = 0
failed_tests = []
for test_category, results in all_results.items():
print(f"\n{test_category.upper()}:")
for test_name, result in results.items():
total_tests += 1
if result == "PASS":
passed_tests += 1
print(f" β {test_name}")
else:
failed_tests.append(f"{test_category}.{test_name}: {result}")
print(f" β {test_name}: {result}")
print(f"\n" + "=" * 60)
print(f"OVERALL RESULTS:")
print(f"Total tests: {total_tests}")
print(f"Passed: {passed_tests}")
print(f"Failed: {len(failed_tests)}")
print(f"Success rate: {passed_tests / total_tests * 100:.1f}%")
if failed_tests:
print(f"\nFAILED TESTS:")
for failure in failed_tests:
print(f" - {failure}")
return len(failed_tests) == 0
def main():
"""Main test function"""
print("TRANSFORMS.PY COMPREHENSIVE TEST SUITE")
print("=" * 60)
print("Testing all functions and classes in transforms.py")
all_results = {}
try:
# Run all test categories
all_results['basic_transforms'] = test_basic_transforms()
all_results['model_specific'] = test_model_specific_transforms()
all_results['advanced_augmentations'] = test_advanced_augmentations()
all_results['advanced_transforms'] = test_advanced_train_transforms()
all_results['denormalize'] = test_denormalize()
all_results['performance'] = test_performance()
all_results['edge_cases'] = test_edge_cases()
# Print final summary
success = print_final_summary(all_results)
if success:
print("\nπ ALL TESTS PASSED! transforms.py is working correctly.")
return 0
else:
print("\nβ SOME TESTS FAILED! Check the issues above.")
return 1
except Exception as e:
print(f"\nπ₯ CRITICAL ERROR during testing: {e}")
print(f"Traceback:\n{traceback.format_exc()}")
return 1
if __name__ == "__main__":
exit_code = main()
print(f"\nTest completed with exit code: {exit_code}") |