Spaces:
Sleeping
Sleeping
File size: 10,941 Bytes
8e1643b | 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 | """
SQLAlchemy models for PDF storage and prediction tracking.
"""
from datetime import datetime
from typing import Dict, Any, List
import json
import pickle
import numpy as np
from sqlalchemy import (
Column, Integer, String, Float, DateTime, Boolean,
Text, LargeBinary, ForeignKey, Index
)
from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import relationship
Base = declarative_base()
class PDFSnapshot(Base):
"""
Stores historical PDF snapshots with all associated data.
Each snapshot represents the option-implied probability distribution
at a specific point in time for a specific expiration.
"""
__tablename__ = 'pdf_snapshots'
# Primary key
id = Column(Integer, primary_key=True, autoincrement=True)
# Timestamp and identification
timestamp = Column(DateTime, nullable=False, index=True)
ticker = Column(String(10), nullable=False, index=True)
# Market data
spot_price = Column(Float, nullable=False)
days_to_expiry = Column(Integer, nullable=False)
expiration_date = Column(DateTime, nullable=False)
# Risk-free rate
risk_free_rate = Column(Float, nullable=False)
# PDF data (stored as binary)
strikes = Column(LargeBinary, nullable=False) # NumPy array pickled
pdf_values = Column(LargeBinary, nullable=False) # NumPy array pickled
# SABR parameters (if used)
sabr_alpha = Column(Float, nullable=True)
sabr_rho = Column(Float, nullable=True)
sabr_nu = Column(Float, nullable=True)
sabr_beta = Column(Float, nullable=True)
interpolation_method = Column(String(20), nullable=True) # 'sabr' or 'spline'
# Statistics (stored as JSON)
statistics = Column(Text, nullable=False) # JSON string
# AI interpretation
interpretation = Column(Text, nullable=True)
interpretation_mode = Column(String(20), nullable=True) # 'standard', 'conservative', etc.
model_used = Column(String(50), nullable=True) # 'ollama' or 'fallback'
# Relationships
pattern_matches = relationship('PatternMatch',
foreign_keys='PatternMatch.current_snapshot_id',
back_populates='current_snapshot')
predictions = relationship('Prediction', back_populates='snapshot')
# Indexes for common queries
__table_args__ = (
Index('idx_ticker_timestamp', 'ticker', 'timestamp'),
Index('idx_ticker_expiry', 'ticker', 'days_to_expiry'),
)
def __repr__(self):
return f"<PDFSnapshot(id={self.id}, ticker={self.ticker}, timestamp={self.timestamp}, dte={self.days_to_expiry})>"
def get_strikes(self) -> np.ndarray:
"""Deserialize strikes from binary."""
return pickle.loads(self.strikes)
def set_strikes(self, strikes: np.ndarray):
"""Serialize strikes to binary."""
self.strikes = pickle.dumps(strikes)
def get_pdf_values(self) -> np.ndarray:
"""Deserialize PDF values from binary."""
return pickle.loads(self.pdf_values)
def set_pdf_values(self, pdf_values: np.ndarray):
"""Serialize PDF values to binary."""
self.pdf_values = pickle.dumps(pdf_values)
def get_statistics(self) -> Dict[str, Any]:
"""Deserialize statistics from JSON."""
return json.loads(self.statistics)
def set_statistics(self, stats: Dict[str, Any]):
"""Serialize statistics to JSON."""
self.statistics = json.dumps(stats)
def to_dict(self) -> Dict[str, Any]:
"""Convert snapshot to dictionary."""
return {
'id': self.id,
'timestamp': self.timestamp.isoformat(),
'ticker': self.ticker,
'spot_price': self.spot_price,
'days_to_expiry': self.days_to_expiry,
'expiration_date': self.expiration_date.isoformat(),
'risk_free_rate': self.risk_free_rate,
'strikes': self.get_strikes().tolist(),
'pdf_values': self.get_pdf_values().tolist(),
'sabr_params': {
'alpha': self.sabr_alpha,
'rho': self.sabr_rho,
'nu': self.sabr_nu,
'beta': self.sabr_beta,
},
'interpolation_method': self.interpolation_method,
'statistics': self.get_statistics(),
'interpretation': self.interpretation,
'interpretation_mode': self.interpretation_mode,
'model_used': self.model_used,
}
class Prediction(Base):
"""
Tracks predictions made from PDFs and their outcomes.
Used to evaluate the accuracy of option-implied probabilities
by comparing forecasted probabilities with actual outcomes.
"""
__tablename__ = 'predictions'
# Primary key
id = Column(Integer, primary_key=True, autoincrement=True)
# Link to PDF snapshot
snapshot_id = Column(Integer, ForeignKey('pdf_snapshots.id'), nullable=False)
# Forecast details
forecast_date = Column(DateTime, nullable=False, index=True)
target_date = Column(DateTime, nullable=False, index=True)
ticker = Column(String(10), nullable=False)
# Prediction
condition = Column(String(20), nullable=False) # 'above', 'below', 'between'
target_level = Column(Float, nullable=False) # Strike or price level
target_level_upper = Column(Float, nullable=True) # For 'between' condition
predicted_probability = Column(Float, nullable=False) # 0.0 to 1.0
# Actual outcome
evaluation_date = Column(DateTime, nullable=True) # When we evaluated
actual_price = Column(Float, nullable=True) # Actual price at target_date
actual_outcome = Column(Boolean, nullable=True) # True if condition met
# Accuracy metrics
accuracy_score = Column(Float, nullable=True) # Brier score or similar
# Additional context
notes = Column(Text, nullable=True)
# Relationship
snapshot = relationship('PDFSnapshot', back_populates='predictions')
# Indexes
__table_args__ = (
Index('idx_ticker_target_date', 'ticker', 'target_date'),
)
def __repr__(self):
return f"<Prediction(id={self.id}, ticker={self.ticker}, target_date={self.target_date}, prob={self.predicted_probability:.2%})>"
def calculate_brier_score(self) -> float:
"""
Calculate Brier score for this prediction.
Brier Score = (predicted_prob - actual_outcome)^2
Lower is better (0 = perfect forecast)
"""
if self.actual_outcome is None:
return None
actual = 1.0 if self.actual_outcome else 0.0
brier = (self.predicted_probability - actual) ** 2
self.accuracy_score = brier
return brier
def to_dict(self) -> Dict[str, Any]:
"""Convert prediction to dictionary."""
return {
'id': self.id,
'snapshot_id': self.snapshot_id,
'forecast_date': self.forecast_date.isoformat(),
'target_date': self.target_date.isoformat(),
'ticker': self.ticker,
'condition': self.condition,
'target_level': self.target_level,
'target_level_upper': self.target_level_upper,
'predicted_probability': self.predicted_probability,
'evaluation_date': self.evaluation_date.isoformat() if self.evaluation_date else None,
'actual_price': self.actual_price,
'actual_outcome': self.actual_outcome,
'accuracy_score': self.accuracy_score,
'notes': self.notes,
}
class PatternMatch(Base):
"""
Stores pattern matching results between current and historical PDFs.
Used to find similar market conditions and compare outcomes.
"""
__tablename__ = 'pattern_matches'
# Primary key
id = Column(Integer, primary_key=True, autoincrement=True)
# Current snapshot being analyzed
current_snapshot_id = Column(Integer, ForeignKey('pdf_snapshots.id'), nullable=False)
# Historical snapshot that matched
historical_snapshot_id = Column(Integer, ForeignKey('pdf_snapshots.id'), nullable=False)
# Match timestamp
match_timestamp = Column(DateTime, nullable=False, default=datetime.utcnow)
# Similarity scores
overall_similarity = Column(Float, nullable=False) # Combined score
shape_similarity = Column(Float, nullable=False) # Cosine similarity
stats_similarity = Column(Float, nullable=False) # Statistical features similarity
# Match rank (1 = best match)
match_rank = Column(Integer, nullable=False)
# Description of historical pattern
description = Column(Text, nullable=True)
# Relationship
current_snapshot = relationship('PDFSnapshot',
foreign_keys=[current_snapshot_id],
back_populates='pattern_matches')
historical_snapshot = relationship('PDFSnapshot', foreign_keys=[historical_snapshot_id])
# Indexes
__table_args__ = (
Index('idx_current_snapshot', 'current_snapshot_id'),
Index('idx_similarity', 'overall_similarity'),
)
def __repr__(self):
return f"<PatternMatch(current={self.current_snapshot_id}, historical={self.historical_snapshot_id}, sim={self.overall_similarity:.2%})>"
def to_dict(self) -> Dict[str, Any]:
"""Convert pattern match to dictionary."""
return {
'id': self.id,
'current_snapshot_id': self.current_snapshot_id,
'historical_snapshot_id': self.historical_snapshot_id,
'match_timestamp': self.match_timestamp.isoformat(),
'overall_similarity': self.overall_similarity,
'shape_similarity': self.shape_similarity,
'stats_similarity': self.stats_similarity,
'match_rank': self.match_rank,
'description': self.description,
}
if __name__ == "__main__":
# Print schema for documentation
print("=" * 80)
print("PDF VISUALIZER DATABASE SCHEMA")
print("=" * 80)
print("\nTABLE: pdf_snapshots")
print("-" * 80)
print("Stores historical PDF snapshots with all associated data")
print("\nColumns:")
for column in PDFSnapshot.__table__.columns:
print(f" - {column.name}: {column.type} {'(PK)' if column.primary_key else ''}")
print("\n\nTABLE: predictions")
print("-" * 80)
print("Tracks predictions and their outcomes for accuracy evaluation")
print("\nColumns:")
for column in Prediction.__table__.columns:
print(f" - {column.name}: {column.type} {'(PK)' if column.primary_key else ''}")
print("\n\nTABLE: pattern_matches")
print("-" * 80)
print("Stores pattern matching results between PDFs")
print("\nColumns:")
for column in PatternMatch.__table__.columns:
print(f" - {column.name}: {column.type} {'(PK)' if column.primary_key else ''}")
print("\n" + "=" * 80)
|