Skip to content

Options Chain Features (OCF)

Options Chain Features (OCF) is a deterministic, research-grade Python library for transforming raw options chain data into model-ready features.

OCF is designed for quantitative research, risk modeling, and systematic analysis workflows where clarity, reproducibility, and explicit data transformations matter more than black-box abstractions.

The library converts vendor-style inputs (e.g. Bloomberg extracts) into aligned canonical tables and derives structured feature sets without imposing trading logic, forecasting assumptions, or execution strategies.

What Problems OCF Solves

Working with options data typically involves:

  • Inconsistent vendor schemas
  • Implicit joins and hidden assumptions
  • Hard-to-reproduce feature pipelines
  • Tight coupling between data, models, and strategy logic

OCF addresses these issues by providing a schema-first, modular pipeline where every transformation is explicit, testable, and independently reusable.

Core Capabilities

OCF provides:

  • Canonical schemas for underlying assets and option chains
  • Normalization layers for raw vendor inputs
  • Deterministic feature blocks operating on aligned daily data
  • Implied volatility surface, skew, and term-structure features
  • Liquidity and positioning features
  • Per-option Black–Scholes Greeks
  • Aggregate exposure representations for ML models
  • End-to-end pipelines with strict validation and configuration

All components can be used independently or composed into a full pipeline.

Design Philosophy

OCF follows a few core principles:

  • No hidden state
    Every transformation is explicit and deterministic.

  • No implicit IO or joins
    File handling and computation are cleanly separated.

  • Schema-driven architecture
    Each stage operates on clearly defined canonical tables.

  • Feature isolation
    Feature blocks can be enabled, disabled, and tested independently.

How to Navigate the Documentation

  • Getting Started -
    High-level overview of inputs, outputs, and workflows.

  • Pipelines -
    How raw data flows through normalization, alignment, feature engineering, and Greeks computation.

  • Features -
    Detailed documentation for each feature block: OHLCV, implied volatility, skew & smile, term structure, and liquidity.

  • Greeks -
    Per-option Greeks, exposure aggregation, and ML-friendly representations.

  • API Reference -
    Stable public entry points for features and Greeks.

  • Schemas & IO -
    Canonical data definitions and supported storage layers.

Project Status

  • Current version: v1.0.0
  • Stability: API-stable for research and production pipelines
  • License: MIT

Source Code

The full source code is available on GitHub:

https://github.com/vansh0016/options-chain-features