File size: 21,343 Bytes
2ed8996 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 | """
Analytics Tests for AegisLM Multi-Run Analysis.
Tests comparison engine, trend analyzer, aggregation utilities,
and error handling for analytics operations.
"""
import pytest
import asyncio
from unittest.mock import AsyncMock, MagicMock, patch
from datetime import datetime, timedelta
from uuid import uuid4
from analytics.comparison_engine import ComparisonEngine, ComparisonMetric, TrendDirection
from analytics.trend_analyzer import TrendAnalyzer, TrendDirection as TrendDir, TrendStrength
from analytics.aggregation_utils import AggregationUtils, AggregationMethod
from analytics.analytics_service import AnalyticsService
from schemas.experiment_schema import Experiment, ExperimentStatus, ResultSummary, ConfigSnapshot
class TestComparisonEngine:
"""Test comparison engine functionality."""
@pytest.fixture
async def comparison_engine(self):
"""Create comparison engine for testing."""
return ComparisonEngine()
@pytest.fixture
def sample_experiments(self):
"""Create sample experiments for testing."""
experiments = []
for i in range(3):
experiment = Experiment(
run_id=uuid4(),
experiment_name=f"Test Experiment {i+1}",
config_snapshot=ConfigSnapshot(
model_name="test-model",
model_config={},
attack_types=["jailbreak"],
pipeline_config={},
prompt_count=10,
max_iterations=5,
mutation_enabled=False,
weights={}
),
model_name="test-model",
dataset_name="test-dataset",
attack_types=["jailbreak"],
prompt_count=10,
status=ExperimentStatus.COMPLETED,
created_at=datetime.utcnow() - timedelta(hours=i),
completed_at=datetime.utcnow() - timedelta(hours=i-1),
result_summary=ResultSummary(
robustness_score=0.5 + (i * 0.1), # 0.5, 0.6, 0.7
risk_score=0.4 - (i * 0.05), # 0.4, 0.35, 0.3
success_rate=0.3 + (i * 0.15), # 0.3, 0.45, 0.6
total_attacks=10,
successful_attacks=3 + i,
failed_attacks=7 - i,
execution_time_ms=1000 + (i * 100)
)
)
experiments.append(experiment)
return experiments
@pytest.mark.asyncio
async def test_compare_runs_success(self, comparison_engine, sample_experiments):
"""Test successful run comparison."""
run_ids = [exp.run_id.hex for exp in sample_experiments]
with patch.object(comparison_engine.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.side_effect = sample_experiments
result = await comparison_engine.compare_runs(run_ids)
assert result is not None
assert len(result.rankings) == 3
assert result.best_run is not None
assert result.worst_run is not None
assert result.total_runs == 3
# Check rankings (higher robustness should be better)
best_run = next(r for r in result.rankings if r.is_best)
worst_run = next(r for r in result.rankings if r.is_worst)
assert best_run.rank == 1
assert worst_run.rank == 3
assert best_run.robustness_score > worst_run.robustness_score
@pytest.mark.asyncio
async def test_compare_runs_insufficient_runs(self, comparison_engine):
"""Test comparison with insufficient runs."""
with pytest.raises(ValueError, match="At least 2 runs required"):
await comparison_engine.compare_runs(["run1"])
@pytest.mark.asyncio
async def test_compare_runs_invalid_run_ids(self, comparison_engine):
"""Test comparison with invalid run IDs."""
with patch.object(comparison_engine.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.return_value = None
with pytest.raises(ValueError, match="Insufficient valid experiments"):
await comparison_engine.compare_runs(["invalid1", "invalid2"])
@pytest.mark.asyncio
async def test_metric_deltas_calculation(self, comparison_engine, sample_experiments):
"""Test metric delta calculations."""
run_ids = [exp.run_id.hex for exp in sample_experiments]
with patch.object(comparison_engine.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.side_effect = sample_experiments
result = await comparison_engine.compare_runs(run_ids)
# Check that deltas are calculated
best_run = next(r for r in result.rankings if r.is_best)
assert len(best_run.deltas_to_best) == 0 # Best run has no deltas to best
assert len(best_run.deltas_to_worst) > 0 # But has deltas to worst
# Check delta structure
for metric_name, delta in best_run.deltas_to_worst.items():
assert delta.metric.value == metric_name
assert isinstance(delta.delta, float)
assert isinstance(delta.delta_percent, float)
assert isinstance(delta.improvement, bool)
class TestTrendAnalyzer:
"""Test trend analyzer functionality."""
@pytest.fixture
async def trend_analyzer(self):
"""Create trend analyzer for testing."""
return TrendAnalyzer()
@pytest.fixture
def sample_trend_experiments(self):
"""Create sample experiments with trend data."""
experiments = []
for i in range(5):
experiment = Experiment(
run_id=uuid4(),
experiment_name=f"Trend Experiment {i+1}",
config_snapshot=ConfigSnapshot(
model_name="trend-model",
model_config={},
attack_types=["jailbreak"],
pipeline_config={},
prompt_count=10,
max_iterations=5,
mutation_enabled=False,
weights={}
),
model_name="trend-model",
dataset_name="trend-dataset",
attack_types=["jailbreak"],
prompt_count=10,
status=ExperimentStatus.COMPLETED,
created_at=datetime.utcnow() - timedelta(days=4-i), # 4,3,2,1,0 days ago
completed_at=datetime.utcnow() - timedelta(days=4-i, hours=-1),
result_summary=ResultSummary(
robustness_score=0.3 + (i * 0.1), # Increasing trend: 0.3, 0.4, 0.5, 0.6, 0.7
risk_score=0.6 - (i * 0.05), # Decreasing trend: 0.6, 0.55, 0.5, 0.45, 0.4
success_rate=0.2 + (i * 0.1), # Increasing trend: 0.2, 0.3, 0.4, 0.5, 0.6
total_attacks=10,
successful_attacks=2 + i,
failed_attacks=8 - i,
execution_time_ms=1000 - (i * 50) # Decreasing trend
)
)
experiments.append(experiment)
return experiments
@pytest.mark.asyncio
async def test_analyze_trend_success(self, trend_analyzer, sample_trend_experiments):
"""Test successful trend analysis."""
run_ids = [exp.run_id.hex for exp in sample_trend_experiments]
with patch.object(trend_analyzer.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.side_effect = sample_trend_experiments
result = await trend_analyzer.analyze_trend(run_ids)
assert result is not None
assert len(result.metric_trends) > 0
assert result.total_runs == 5
assert result.time_period_days >= 4
assert result.overall_direction in [TrendDir.INCREASING, TrendDir.DECREASING, TrendDir.STABLE]
assert 0 <= result.overall_health_score <= 1
@pytest.mark.asyncio
async def test_analyze_trend_insufficient_runs(self, trend_analyzer):
"""Test trend analysis with insufficient runs."""
with pytest.raises(ValueError, match="At least 3 runs required"):
await trend_analyzer.analyze_trend(["run1", "run2"])
@pytest.mark.asyncio
async def test_trend_direction_detection(self, trend_analyzer, sample_trend_experiments):
"""Test trend direction detection."""
run_ids = [exp.run_id.hex for exp in sample_trend_experiments]
with patch.object(trend_analyzer.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.side_effect = sample_trend_experiments
result = await trend_analyzer.analyze_trend(run_ids)
# Check robustness trend (should be increasing)
robustness_trend = result.metric_trends.get('robustness_score')
if robustness_trend:
assert robustness_trend.metrics.direction in [TrendDir.INCREASING, TrendDir.STABLE]
assert robustness_trend.metrics.strength in [TrendStrength.WEAK, TrendStrength.MODERATE, TrendStrength.STRONG]
@pytest.mark.asyncio
async def test_anomaly_detection(self, trend_analyzer):
"""Test anomaly detection in trends."""
# Create experiments with an anomaly
experiments = []
base_value = 0.5
for i in range(5):
value = base_value + (i * 0.1)
# Add anomaly in the middle
if i == 2:
value = 0.9 # Much higher than expected
experiment = Experiment(
run_id=uuid4(),
experiment_name=f"Anomaly Test {i+1}",
config_snapshot=ConfigSnapshot(
model_name="test-model",
model_config={},
attack_types=["jailbreak"],
pipeline_config={},
prompt_count=10,
max_iterations=5,
mutation_enabled=False,
weights={}
),
model_name="test-model",
dataset_name="test-dataset",
attack_types=["jailbreak"],
prompt_count=10,
status=ExperimentStatus.COMPLETED,
created_at=datetime.utcnow() - timedelta(days=4-i),
result_summary=ResultSummary(
robustness_score=value,
risk_score=0.3,
success_rate=0.5,
total_attacks=10,
successful_attacks=5,
failed_attacks=5
)
)
experiments.append(experiment)
run_ids = [exp.run_id.hex for exp in experiments]
with patch.object(trend_analyzer.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.side_effect = experiments
result = await trend_analyzer.analyze_trend(run_ids)
# Check if anomaly was detected
robustness_trend = result.metric_trends.get('robustness_score')
if robustness_trend:
# Should detect at least one anomaly due to the outlier
assert robustness_trend.metrics.stability_score < 0.9 # Lower stability due to anomaly
class TestAggregationUtils:
"""Test aggregation utilities functionality."""
@pytest.fixture
async def aggregation_utils(self):
"""Create aggregation utilities for testing."""
return AggregationUtils()
@pytest.fixture
def sample_aggregation_experiments(self):
"""Create sample experiments for aggregation testing."""
experiments = []
models = ["model-a", "model-b", "model-a"]
datasets = ["dataset-1", "dataset-1", "dataset-2"]
for i in range(3):
experiment = Experiment(
run_id=uuid4(),
experiment_name=f"Aggregation Test {i+1}",
config_snapshot=ConfigSnapshot(
model_name=models[i],
model_config={},
attack_types=["jailbreak"],
pipeline_config={},
prompt_count=10,
max_iterations=5,
mutation_enabled=False,
weights={}
),
model_name=models[i],
dataset_name=datasets[i],
attack_types=["jailbreak"],
prompt_count=10,
status=ExperimentStatus.COMPLETED,
created_at=datetime.utcnow() - timedelta(hours=i),
result_summary=ResultSummary(
robustness_score=0.4 + (i * 0.1),
risk_score=0.5 - (i * 0.05),
success_rate=0.3 + (i * 0.2),
total_attacks=10,
successful_attacks=3 + i,
failed_attacks=7 - i,
execution_time_ms=1000 + (i * 200)
)
)
experiments.append(experiment)
return experiments
@pytest.mark.asyncio
async def test_aggregate_metrics_success(self, aggregation_utils, sample_aggregation_experiments):
"""Test successful metrics aggregation."""
result = await aggregation_utils.aggregate_metrics(sample_aggregation_experiments)
assert result is not None
assert result.total_experiments == 3
assert result.completed_experiments == 3
assert result.success_rate == 1.0
assert result.overall_health_score >= 0
assert result.model_distribution == {"model-a": 2, "model-b": 1}
assert result.dataset_distribution == {"dataset-1": 2, "dataset-2": 1}
@pytest.mark.asyncio
async def test_aggregate_empty_experiments(self, aggregation_utils):
"""Test aggregation with empty experiments list."""
result = await aggregation_utils.aggregate_metrics([])
assert result.total_experiments == 0
assert result.completed_experiments == 0
assert result.success_rate == 0.0
@pytest.mark.asyncio
async def test_aggregate_by_model(self, aggregation_utils, sample_aggregation_experiments):
"""Test aggregation by model."""
result = await aggregation_utils.aggregate_by_model(sample_aggregation_experiments)
assert "model-a" in result
assert "model-b" in result
assert result["model-a"].total_experiments == 2
assert result["model-b"].total_experiments == 1
@pytest.mark.asyncio
async def test_aggregate_by_dataset(self, aggregation_utils, sample_aggregation_experiments):
"""Test aggregation by dataset."""
result = await aggregation_utils.aggregate_by_dataset(sample_aggregation_experiments)
assert "dataset-1" in result
assert "dataset-2" in result
assert result["dataset-1"].total_experiments == 2
assert result["dataset-2"].total_experiments == 1
@pytest.mark.asyncio
async def test_get_top_performers(self, aggregation_utils, sample_aggregation_experiments):
"""Test getting top performers."""
top_performers = await aggregation_utils.get_top_performers(
sample_aggregation_experiments,
"robustness_score",
3
)
assert len(top_performers) == 3
assert top_performers[0][1] >= top_performers[1][1] # Should be sorted descending
assert top_performers[1][1] >= top_performers[2][1]
class TestAnalyticsService:
"""Test analytics service functionality."""
@pytest.fixture
async def analytics_service(self):
"""Create analytics service for testing."""
mock_db = AsyncMock()
return AnalyticsService(mock_db)
@pytest.mark.asyncio
async def test_fetch_runs_by_ids_success(self, analytics_service, sample_experiments):
"""Test successful run fetching by IDs."""
run_ids = [exp.run_id.hex for exp in sample_experiments[:2]]
with patch.object(analytics_service.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.side_effect = sample_experiments[:2]
result = await analytics_service.fetch_runs_by_ids(run_ids)
assert len(result) == 2
assert all(exp.run_id.hex in run_ids for exp in result)
@pytest.mark.asyncio
async def test_fetch_runs_by_ids_not_found(self, analytics_service):
"""Test fetching non-existent runs."""
with patch.object(analytics_service.experiment_manager.store, 'get_experiment') as mock_get:
mock_get.return_value = None
with pytest.raises(Exception): # Should raise HTTPException
await analytics_service.fetch_runs_by_ids(["non-existent"])
@pytest.mark.asyncio
async def test_compare_runs_integration(self, analytics_service, sample_experiments):
"""Test compare runs integration."""
run_ids = [exp.run_id.hex for exp in sample_experiments]
with patch.object(analytics_service, 'fetch_runs_by_ids') as mock_fetch:
mock_fetch.return_value = sample_experiments
with patch.object(analytics_service.comparison_engine, 'compare_runs') as mock_compare:
mock_compare.return_value = MagicMock()
result = await analytics_service.compare_runs(run_ids)
mock_fetch.assert_called_once_with(run_ids, None)
mock_compare.assert_called_once_with(run_ids)
@pytest.mark.asyncio
async def test_analyze_trends_integration(self, analytics_service, sample_experiments):
"""Test trend analysis integration."""
run_ids = [exp.run_id.hex for exp in sample_experiments]
with patch.object(analytics_service, 'fetch_runs_by_ids') as mock_fetch:
mock_fetch.return_value = sample_experiments
with patch.object(analytics_service.trend_analyzer, 'analyze_trend') as mock_analyze:
mock_analyze.return_value = MagicMock()
result = await analytics_service.analyze_trends(run_ids)
mock_fetch.assert_called_once_with(run_ids, None)
mock_analyze.assert_called_once_with(run_ids)
class TestErrorHandling:
"""Test error handling in analytics operations."""
@pytest.mark.asyncio
async def test_comparison_engine_error_handling(self):
"""Test comparison engine error handling."""
engine = ComparisonEngine()
# Test with invalid UUID format
with pytest.raises(ValueError):
await engine.compare_runs(["invalid-uuid"])
# Test with insufficient runs
with pytest.raises(ValueError, match="At least 2 runs required"):
await engine.compare_runs([])
@pytest.mark.asyncio
async def test_trend_analyzer_error_handling(self):
"""Test trend analyzer error handling."""
analyzer = TrendAnalyzer()
# Test with insufficient runs
with pytest.raises(ValueError, match="At least 3 runs required"):
await analyzer.analyze_trend(["run1", "run2"])
# Test with invalid UUID format
with pytest.raises(ValueError):
await analyzer.analyze_trend(["invalid-uuid", "run2", "run3"])
@pytest.mark.asyncio
async def test_aggregation_utils_error_handling(self):
"""Test aggregation utilities error handling."""
utils = AggregationUtils()
# Test with empty values
with pytest.raises(ValueError):
await utils._calculate_metric_statistics("test_metric", [])
def test_metric_validation(self):
"""Test metric validation and edge cases."""
from analytics.comparison_engine import ComparisonMetric
# Test valid metrics
assert ComparisonMetric.ROBUSTNESS_SCORE == "robustness_score"
assert ComparisonMetric.RISK_SCORE == "risk_score"
# Test improvement logic
from analytics.comparison_engine import ComparisonEngine
engine = ComparisonEngine()
# Higher robustness should be improvement
assert engine._is_improvement(ComparisonMetric.ROBUSTNESS_SCORE, 0.1) is True
assert engine._is_improvement(ComparisonMetric.ROBUSTNESS_SCORE, -0.1) is False
# Lower risk should be improvement
assert engine._is_improvement(ComparisonMetric.RISK_SCORE, -0.1) is True
assert engine._is_improvement(ComparisonMetric.RISK_SCORE, 0.1) is False
if __name__ == "__main__":
pytest.main([__file__, "-v"])
|