Spaces:
Sleeping
Sleeping
File size: 10,167 Bytes
68b32d7 | 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 | """
A/B Testing API Endpoints
REST API for creating and managing A/B tests for agent configuration,
prompt, strategy, and tool comparisons.
Endpoints:
- POST /api/ab-tests/create - Create new A/B test
- POST /api/ab-tests/{test_id}/start - Start a test
- POST /api/ab-tests/{test_id}/complete - Complete a test and get results
- POST /api/ab-tests/{test_id}/assign - Assign user to variant
- POST /api/ab-tests/{test_id}/record - Record metric for participant
- GET /api/ab-tests/{test_id}/results - Get test results
- GET /api/ab-tests - List all tests
"""
from datetime import datetime
import logging
from typing import Any, Dict, List, Optional
from fastapi import Depends, Query
from pydantic import BaseModel, Field
from sqlalchemy.orm import Session
from core.ab_testing_service import ABTestingService
from core.base_routes import BaseAPIRouter
from core.database import get_db
logger = logging.getLogger(__name__)
router = BaseAPIRouter(prefix="/api/ab-tests", tags=["A/B Testing"])
# ========================================================================
# Request/Response Models
# ========================================================================
class CreateTestRequest(BaseModel):
"""Request to create a new A/B test."""
name: str = Field(..., description="Test name")
test_type: str = Field(..., description="Type of test (agent_config, prompt, strategy, tool)")
agent_id: str = Field(..., description="ID of agent to test")
variant_a_config: Dict[str, Any] = Field(..., description="Configuration for control variant")
variant_b_config: Dict[str, Any] = Field(..., description="Configuration for treatment variant")
primary_metric: str = Field(..., description="Primary success metric")
variant_a_name: str = Field(default="Control", description="Name for variant A")
variant_b_name: str = Field(default="Treatment", description="Name for variant B")
description: Optional[str] = Field(None, description="Test description")
traffic_percentage: float = Field(default=0.5, ge=0.0, le=1.0, description="Traffic to variant B")
min_sample_size: int = Field(default=100, ge=1, description="Min sample size per variant")
confidence_level: float = Field(default=0.95, ge=0.0, le=1.0, description="Confidence level")
secondary_metrics: Optional[List[str]] = Field(default=None, description="Additional metrics")
class AssignVariantRequest(BaseModel):
"""Request to assign user to variant."""
user_id: str = Field(..., description="User ID")
session_id: Optional[str] = Field(None, description="Session ID")
class RecordMetricRequest(BaseModel):
"""Request to record metric for participant."""
user_id: str = Field(..., description="User ID")
success: Optional[bool] = Field(None, description="Boolean success indicator")
metric_value: Optional[float] = Field(None, description="Numerical metric value")
metadata: Optional[Dict[str, Any]] = Field(default=None, description="Additional metadata")
# ========================================================================
# Test Management Endpoints
# ========================================================================
@router.post("/create")
async def create_test(
request: CreateTestRequest,
db: Session = Depends(get_db)
):
"""
Create a new A/B test.
Tests different agent configurations, prompts, or strategies
to measure impact on key metrics.
Request Body:
- name: Test name
- test_type: Type (agent_config, prompt, strategy, tool)
- agent_id: Agent to test
- variant_a_config: Control configuration
- variant_b_config: Treatment configuration
- primary_metric: Success metric (satisfaction_rate, success_rate, response_time)
- traffic_percentage: Fraction to variant B (default: 0.5)
- min_sample_size: Min sample size per variant (default: 100)
- confidence_level: Statistical confidence (default: 0.95)
Response:
Created test data with test_id
"""
service = ABTestingService(db)
result = service.create_test(
name=request.name,
test_type=request.test_type,
agent_id=request.agent_id,
variant_a_config=request.variant_a_config,
variant_b_config=request.variant_b_config,
primary_metric=request.primary_metric,
variant_a_name=request.variant_a_name,
variant_b_name=request.variant_b_name,
description=request.description,
traffic_percentage=request.traffic_percentage,
min_sample_size=request.min_sample_size,
confidence_level=request.confidence_level,
secondary_metrics=request.secondary_metrics
)
if "error" in result:
raise router.error_response(
error_code="AB_TEST_ERROR",
message=result["error"],
status_code=400
)
return router.success_response(data=result)
@router.post("/{test_id}/start")
async def start_test(
test_id: str,
db: Session = Depends(get_db)
):
"""
Start an A/B test.
Changes test status from 'draft' to 'running' and
begins variant assignment.
Response:
Updated test data with started_at timestamp
"""
service = ABTestingService(db)
result = service.start_test(test_id)
if "error" in result:
raise router.error_response(
error_code="AB_TEST_ERROR",
message=result["error"],
status_code=400
)
return router.success_response(data=result)
@router.post("/{test_id}/complete")
async def complete_test(
test_id: str,
db: Session = Depends(get_db)
):
"""
Complete an A/B test and calculate results.
Performs statistical analysis to determine if there's
a significant difference between variants.
Response:
Test results including:
- variant_a_metrics: Metrics for control
- variant_b_metrics: Metrics for treatment
- p_value: Statistical significance
- winner: 'A', 'B', or 'inconclusive'
"""
service = ABTestingService(db)
result = service.complete_test(test_id)
if "error" in result:
raise router.error_response(
error_code="AB_TEST_ERROR",
message=result["error"],
status_code=400
)
return router.success_response(data=result)
# ========================================================================
# Variant Assignment Endpoints
# ========================================================================
@router.post("/{test_id}/assign")
async def assign_variant(
test_id: str,
request: AssignVariantRequest,
db: Session = Depends(get_db)
):
"""
Assign a user to a test variant.
Uses deterministic hash-based assignment to ensure
consistent assignment for the same user.
Request Body:
- user_id: User ID
- session_id: Optional session ID
Response:
Assignment data with:
- variant: 'A' or 'B'
- variant_name: Human-readable variant name
- config: Variant configuration
- existing_assignment: Boolean
"""
service = ABTestingService(db)
result = service.assign_variant(
test_id=test_id,
user_id=request.user_id,
session_id=request.session_id
)
if "error" in result:
raise router.error_response(
error_code="AB_TEST_ERROR",
message=result["error"],
status_code=400
)
return router.success_response(data=result)
@router.post("/{test_id}/record")
async def record_metric(
test_id: str,
request: RecordMetricRequest,
db: Session = Depends(get_db)
):
"""
Record a metric for a test participant.
Tracks outcome data for statistical analysis.
Request Body:
- user_id: User ID
- success: Boolean success (optional)
- metric_value: Numerical value (optional)
- metadata: Additional data (optional)
Response:
Recorded metric data
"""
service = ABTestingService(db)
result = service.record_metric(
test_id=test_id,
user_id=request.user_id,
success=request.success,
metric_value=request.metric_value,
metadata=request.metadata
)
if "error" in result:
raise router.error_response(
error_code="AB_TEST_ERROR",
message=result["error"],
status_code=400
)
return router.success_response(data=result)
# ========================================================================
# Results and Analytics Endpoints
# ========================================================================
@router.get("/{test_id}/results")
async def get_test_results(
test_id: str,
db: Session = Depends(get_db)
):
"""
Get current results for an A/B test.
Returns participant counts and metrics for both variants.
Response:
Test results with:
- variant_a: Control variant data
- variant_b: Treatment variant data
- winner: Test winner (if completed)
- statistical_significance: p-value
"""
service = ABTestingService(db)
result = service.get_test_results(test_id)
if "error" in result:
raise router.not_found_error("ABTest", test_id, details={"error": result["error"]})
return router.success_response(data=result)
@router.get("")
async def list_tests(
agent_id: Optional[str] = Query(None, description="Filter by agent ID"),
status: Optional[str] = Query(None, description="Filter by status"),
limit: int = Query(50, ge=1, le=100, description="Max results"),
db: Session = Depends(get_db)
):
"""
List A/B tests with optional filtering.
Query Parameters:
- agent_id: Optional agent filter
- status: Optional status filter (draft, running, paused, completed)
- limit: Maximum results (default: 50)
Response:
List of tests with summary data
"""
service = ABTestingService(db)
return service.list_tests(
agent_id=agent_id,
status=status,
limit=limit
)
|