Spaces:
Runtime error
Runtime error
File size: 15,362 Bytes
8366ee8 | 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 | """
Evaluation Reporting
====================
Generate comprehensive evaluation reports for model performance and predictions.
"""
import numpy as np
import pandas as pd
from typing import Dict, List, Optional, Any
from dataclasses import dataclass
from datetime import datetime, timedelta
import logging
import json
from pathlib import Path
logger = logging.getLogger(__name__)
@dataclass
class ReportConfig:
"""Configuration for report generation."""
output_dir: str = "reports"
include_charts: bool = True
format: str = "markdown" # 'markdown', 'html', 'json'
date_range_days: int = 30
class EvaluationReporter:
"""
Generates comprehensive evaluation reports.
Reports include:
- Model accuracy metrics
- ROI and profit analysis
- Market-specific performance
- Time-series analysis
- Confidence calibration
"""
def __init__(self, config: ReportConfig = None):
self.config = config or ReportConfig()
self.output_dir = Path(self.config.output_dir)
self.output_dir.mkdir(parents=True, exist_ok=True)
def generate_accuracy_report(
self,
predictions: List[Dict],
results: List[Dict]
) -> Dict[str, Any]:
"""
Generate accuracy metrics report.
Args:
predictions: List of prediction records
results: List of actual results
Returns:
Report dictionary with metrics
"""
if not predictions or not results:
return {'error': 'No data provided'}
# Match predictions to results
matched = self._match_predictions_results(predictions, results)
if not matched:
return {'error': 'No matched predictions'}
# Calculate metrics
correct = sum(1 for m in matched if m['correct'])
total = len(matched)
accuracy = correct / total if total > 0 else 0
# By confidence level
confidence_buckets = self._bucket_by_confidence(matched)
# By market
market_accuracy = self._accuracy_by_market(matched)
# By time period
time_analysis = self._accuracy_over_time(matched)
report = {
'summary': {
'total_predictions': total,
'correct': correct,
'accuracy': round(accuracy * 100, 2),
'generated_at': datetime.now().isoformat()
},
'by_confidence': confidence_buckets,
'by_market': market_accuracy,
'time_analysis': time_analysis
}
return report
def generate_roi_report(
self,
bets: List[Dict]
) -> Dict[str, Any]:
"""
Generate ROI and profit analysis report.
Args:
bets: List of bet records with stakes, odds, and results
Returns:
ROI report
"""
if not bets:
return {'error': 'No bet data'}
total_stake = sum(b.get('stake', 0) for b in bets)
total_returns = sum(
b.get('stake', 0) * b.get('odds', 1)
for b in bets if b.get('won', False)
)
profit = total_returns - total_stake
roi = (profit / total_stake * 100) if total_stake > 0 else 0
# Win rate
wins = sum(1 for b in bets if b.get('won', False))
win_rate = wins / len(bets) * 100 if bets else 0
# By market
market_roi = {}
for b in bets:
market = b.get('market', 'unknown')
if market not in market_roi:
market_roi[market] = {'stake': 0, 'returns': 0, 'bets': 0}
market_roi[market]['stake'] += b.get('stake', 0)
market_roi[market]['bets'] += 1
if b.get('won', False):
market_roi[market]['returns'] += b.get('stake', 0) * b.get('odds', 1)
for market, data in market_roi.items():
data['roi'] = round(
(data['returns'] - data['stake']) / data['stake'] * 100
if data['stake'] > 0 else 0, 2
)
return {
'summary': {
'total_bets': len(bets),
'total_stake': round(total_stake, 2),
'total_returns': round(total_returns, 2),
'profit': round(profit, 2),
'roi': round(roi, 2),
'win_rate': round(win_rate, 2)
},
'by_market': market_roi
}
def generate_model_comparison_report(
self,
model_results: Dict[str, List[Dict]]
) -> Dict[str, Any]:
"""
Compare performance across multiple models.
Args:
model_results: Dict mapping model name to prediction results
Returns:
Comparison report
"""
comparison = {}
for model_name, results in model_results.items():
if not results:
continue
correct = sum(1 for r in results if r.get('correct', False))
total = len(results)
# Brier score
brier_scores = []
for r in results:
prob = r.get('confidence', 0.5)
outcome = 1 if r.get('correct', False) else 0
brier_scores.append((prob - outcome) ** 2)
avg_brier = np.mean(brier_scores) if brier_scores else 0
comparison[model_name] = {
'total': total,
'correct': correct,
'accuracy': round(correct / total * 100, 2) if total > 0 else 0,
'brier_score': round(avg_brier, 4),
'avg_confidence': round(np.mean([r.get('confidence', 0.5) for r in results]), 3)
}
# Rank models
ranked = sorted(
comparison.items(),
key=lambda x: x[1]['accuracy'],
reverse=True
)
return {
'models': comparison,
'ranking': [{'model': m, 'accuracy': d['accuracy']} for m, d in ranked],
'best_model': ranked[0][0] if ranked else None
}
def generate_calibration_report(
self,
predictions: List[Dict]
) -> Dict[str, Any]:
"""
Generate probability calibration report.
"""
buckets = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]
calibration = {f"0.{int(b*10-1)}-{b}": {'count': 0, 'wins': 0} for b in buckets}
for pred in predictions:
conf = pred.get('confidence', 0.5)
correct = pred.get('correct', False)
for b in buckets:
if conf <= b:
key = f"0.{int(b*10-1)}-{b}"
calibration[key]['count'] += 1
if correct:
calibration[key]['wins'] += 1
break
# Calculate actual vs expected
for key, data in calibration.items():
if data['count'] > 0:
data['actual_rate'] = round(data['wins'] / data['count'], 3)
expected = float(key.split('-')[1])
data['expected_rate'] = expected
data['calibration_error'] = round(abs(data['actual_rate'] - expected), 3)
else:
data['actual_rate'] = 0
data['expected_rate'] = float(key.split('-')[1])
data['calibration_error'] = 0
# Overall calibration error
errors = [d['calibration_error'] for d in calibration.values() if d['count'] > 0]
avg_error = np.mean(errors) if errors else 0
return {
'buckets': calibration,
'average_calibration_error': round(avg_error, 4),
'is_well_calibrated': avg_error < 0.05
}
def save_report(
self,
report: Dict,
filename: str
) -> str:
"""Save report to file."""
filepath = self.output_dir / filename
if self.config.format == 'json':
filepath = filepath.with_suffix('.json')
with open(filepath, 'w') as f:
json.dump(report, f, indent=2, default=str)
elif self.config.format == 'markdown':
filepath = filepath.with_suffix('.md')
md_content = self._report_to_markdown(report)
with open(filepath, 'w') as f:
f.write(md_content)
elif self.config.format == 'html':
filepath = filepath.with_suffix('.html')
html_content = self._report_to_html(report)
with open(filepath, 'w') as f:
f.write(html_content)
logger.info(f"Report saved to {filepath}")
return str(filepath)
def _match_predictions_results(
self,
predictions: List[Dict],
results: List[Dict]
) -> List[Dict]:
"""Match predictions with actual results."""
matched = []
results_map = {
(r.get('home_team'), r.get('away_team'), r.get('date')): r
for r in results
}
for pred in predictions:
key = (pred.get('home_team'), pred.get('away_team'), pred.get('date'))
if key in results_map:
result = results_map[key]
matched.append({
**pred,
'actual_result': result.get('result'),
'correct': pred.get('prediction') == result.get('result')
})
return matched
def _bucket_by_confidence(self, matched: List[Dict]) -> Dict:
"""Bucket accuracy by confidence level."""
buckets = {
'very_high (90%+)': [],
'high (75-90%)': [],
'medium (60-75%)': [],
'low (<60%)': []
}
for m in matched:
conf = m.get('confidence', 0.5)
if conf >= 0.9:
buckets['very_high (90%+)'].append(m)
elif conf >= 0.75:
buckets['high (75-90%)'].append(m)
elif conf >= 0.6:
buckets['medium (60-75%)'].append(m)
else:
buckets['low (<60%)'].append(m)
result = {}
for bucket, items in buckets.items():
if items:
correct = sum(1 for i in items if i['correct'])
result[bucket] = {
'total': len(items),
'correct': correct,
'accuracy': round(correct / len(items) * 100, 2)
}
return result
def _accuracy_by_market(self, matched: List[Dict]) -> Dict:
"""Calculate accuracy by market type."""
markets = {}
for m in matched:
market = m.get('market', 'unknown')
if market not in markets:
markets[market] = {'total': 0, 'correct': 0}
markets[market]['total'] += 1
if m['correct']:
markets[market]['correct'] += 1
for market, data in markets.items():
data['accuracy'] = round(
data['correct'] / data['total'] * 100, 2
) if data['total'] > 0 else 0
return markets
def _accuracy_over_time(self, matched: List[Dict]) -> Dict:
"""Analyze accuracy over time."""
# Group by week
weekly = {}
for m in matched:
date_str = m.get('date', '')
if date_str:
try:
date = datetime.fromisoformat(date_str)
week = date.strftime('%Y-W%W')
if week not in weekly:
weekly[week] = {'total': 0, 'correct': 0}
weekly[week]['total'] += 1
if m['correct']:
weekly[week]['correct'] += 1
except:
pass
for week, data in weekly.items():
data['accuracy'] = round(
data['correct'] / data['total'] * 100, 2
) if data['total'] > 0 else 0
return weekly
def _report_to_markdown(self, report: Dict) -> str:
"""Convert report to markdown format."""
lines = ["# Evaluation Report\n"]
lines.append(f"Generated: {datetime.now().strftime('%Y-%m-%d %H:%M')}\n")
if 'summary' in report:
lines.append("\n## Summary\n")
for key, value in report['summary'].items():
lines.append(f"- **{key}**: {value}")
if 'by_market' in report:
lines.append("\n## Performance by Market\n")
lines.append("| Market | Total | Correct | Accuracy |")
lines.append("|--------|-------|---------|----------|")
for market, data in report['by_market'].items():
lines.append(
f"| {market} | {data.get('total', 0)} | "
f"{data.get('correct', 0)} | {data.get('accuracy', 0)}% |"
)
if 'by_confidence' in report:
lines.append("\n## Performance by Confidence\n")
for bucket, data in report['by_confidence'].items():
lines.append(f"- **{bucket}**: {data.get('accuracy', 0)}% ({data.get('total', 0)} predictions)")
return '\n'.join(lines)
def _report_to_html(self, report: Dict) -> str:
"""Convert report to HTML format."""
html = """
<!DOCTYPE html>
<html>
<head>
<title>Evaluation Report</title>
<style>
body { font-family: Arial, sans-serif; margin: 20px; }
table { border-collapse: collapse; width: 100%; margin: 20px 0; }
th, td { border: 1px solid #ddd; padding: 8px; text-align: left; }
th { background-color: #4CAF50; color: white; }
.metric { font-size: 24px; font-weight: bold; }
</style>
</head>
<body>
<h1>Evaluation Report</h1>
"""
if 'summary' in report:
html += "<h2>Summary</h2><ul>"
for key, value in report['summary'].items():
html += f"<li><strong>{key}</strong>: {value}</li>"
html += "</ul>"
html += "</body></html>"
return html
# Global instance
_reporter: Optional[EvaluationReporter] = None
def get_reporter() -> EvaluationReporter:
"""Get or create reporter instance."""
global _reporter
if _reporter is None:
_reporter = EvaluationReporter()
return _reporter
def generate_report(
predictions: List[Dict],
results: List[Dict],
save: bool = True
) -> Dict:
"""Quick function to generate and optionally save a report."""
reporter = get_reporter()
report = reporter.generate_accuracy_report(predictions, results)
if save:
filename = f"report_{datetime.now().strftime('%Y%m%d_%H%M%S')}"
reporter.save_report(report, filename)
return report
|