Polaris Data Example Notebooks

This workspace packages Jupyter notebooks that demonstrate practical polaris-data usage against live Polaris market data.

The examples are meant to show a few core patterns clearly:

• discovering markets from the Polaris catalog
• fetching bounded OHLCV windows
• fetching raw trade streams
• working with standardized order-book schemas
• comparing bucketed derived metrics like VWAP and realized volatility
• turning API responses into Pandas analysis frames
• building quick visual checks with Matplotlib

Notebook Overview

notebooks/hyperliquid_btc_trade_analysis.ipynb

A compact single-market walkthrough using Hyperliquid BTC data. It shows the most direct end-to-end polaris-data flow:

• discover the market with PolarisClient.catalog(...)
• fetch 1-minute OHLCV bars with PolarisClient.ohlcv(...)
• fetch the matching raw trades with PolarisClient.trades(...)
• summarize buy and sell flow
• visualize price alongside signed trade pressure

This is the best starting point if someone wants to understand the package API shape quickly.

notebooks/lighter_aapl_standardized_schema_tour.ipynb

A second walkthrough focused on the broader standardized schema API surface using the public lighter AAPL perpetual market. It shows how to:

• resolve Lighter's numeric market id from catalog(...)
• select a short recent window with public coverage
• inspect events(...) and l2_snapshots(...)
• track quotes with bbo(...)
• compute bucketed volume(...), vwap(...), and volatility(...)
• handle empty funding_rates(...) and mark_prices(...) responses without assuming coverage

This is the better reference if someone wants to explore order-book-oriented standardized methods beyond the first trade-and-bar example.

Repo Layout

.
├── notebooks/
│   ├── hyperliquid_btc_trade_analysis.ipynb
│   └── lighter_aapl_standardized_schema_tour.ipynb
├── Makefile
├── pyproject.toml
└── uv.lock

Quickstart

This repo uses uv for environment management.

make install
make notebook

Then open the notebooks from the JupyterLab file browser.

Notes

• The notebooks intentionally use explicit bounded time windows so they stay fast to rerun.
• Outputs have been cleared so the examples are portable and do not carry stale environment-specific artifacts.
• If you want to extend the examples, the easiest next step is to widen the ticker basket or swap in different source and market combinations.