Full Limit Order Book

Overview

The Full Limit Order Book (LOB) data type provides a comprehensive view of the order book for a specific trading pair on an exchange. It captures all changes to the order book, allowing for a detailed reconstruction of the market state at any given moment.

File Organization

Full Limit Order Book data is organized in the S3 buckets using the structure below. Paths are identical in coinapi and coinapi-daily-tail; the bucket and partition date determine which dataset you are reading.

T-LIMITBOOK_FULL/
└── D-{YYYYMMDD | YYYYMMDDHH}/
    └── E-[EXCHANGE]/
        └── IDDI-[IDENTIFIER]+SC-[COINAPI_SYMBOL_ID]+S-[EXCHANGE_SYMBOL].csv.gz

Examples:

T-LIMITBOOK_FULL/D-20230701/E-BINANCE/IDDI-1234+SC-BINANCE_SPOT_ETH_USDT+S-ETH-USDT.csv.gz

T-LIMITBOOK_FULL/D-2026011722/E-COINBASE/IDDI-6002426+SC-COINBASE_SPOT_AVAX_USD+S-AVAX__002DUSD.csv.gz

Where:

• T-LIMITBOOK_FULL: Indicates the data type (Full Limit Order Book)
• D-{YYYYMMDD | YYYYMMDDHH} defines the time partition:
  • YYYYMMDD in coinapi → legacy daily partitions published before 2026-06-09
  • YYYYMMDD in coinapi-daily-tail → previous day only, retained 24 hours
  • YYYYMMDDHH in coinapi → hourly partitions for data published on or after 2026-06-09
• E-[EXCHANGE]: Exchange name
• IDDI-[IDENTIFIER]: CoinAPI's internal identifier
• SC-[COINAPI_SYMBOL_ID]: CoinAPI's symbol identifier
• S-[EXCHANGE_SYMBOL]: The symbol as identified by the exchange

File Format

Files are in CSV format, compressed with gzip. Each row represents a single update to the order book.

Data Fields

Example Data

time_exchange,time_coinapi,is_buy,update_type,entry_px,entry_sx,order_id
2023-07-01T12:00:00.123456Z,2023-07-01T12:00:00.234567Z,1,ADD,30000.00,1.5,ord123
2023-07-01T12:00:01.123456Z,2023-07-01T12:00:01.234567Z,0,DELETE,30005.00,0.5,ord124
2023-07-01T12:00:02.123456Z,2023-07-01T12:00:02.234567Z,1,SUB,30000.00,0.5,ord123

Data Collection Process

1. We maintain a real-time connection to each supported exchange.
2. Every change to the order book is captured and recorded.
3. The data is processed, normalized, and stored in our system.
4. Data is uploaded to the coinapi bucket as hourly partitions for data published on or after 2026-06-09. Legacy daily partitions published before that date remain in coinapi. The coinapi-daily-tail bucket receives daily files for the previous day only and retains them for 24 hours.

Update Types Explained

• ADD: Adds a new order to the book.
• SUB: Subtracts volume from an existing order.
• MATCH: Similar to SUB, but indicates the subtraction was due to a trade execution.
• SET: Sets a new volume for an existing order.
• DELETE: Removes an order from the book.
• SNAPSHOT: Provides a full snapshot of the order book. All previous data should be discarded before processing a SNAPSHOT.

When update_type is numeric, use this mapping:

• 0 = ADD
• 1 = SUB
• 2 = SET
• 3 = DELETE
• 4 = MATCH

Note: If you see numeric values in Prediction Markets Flat Files, treat them according to the mapping above.

Corner Cases and Special Considerations

1. Order ID Usage: Some exchanges provide order IDs (Level 3 data), while others don't (Level 2 data). When order IDs are not available, the order_id field will be empty.
2. Snapshots: Snapshots are typically provided at the start of each day or after a disconnection. Always process SNAPSHOT updates by clearing the existing book first.
3. High Frequency Updates: During periods of high volatility, you may observe a very high frequency of updates.

Usage Tips

1. To reconstruct the full order book, process updates sequentially, maintaining the state of the book in memory.
2. Use the time_exchange field for accurate sequencing of updates.
3. Pay attention to the update_type field to correctly apply changes to your local order book representation.

For any questions or issues with the Full Limit Order Book data, please contact our support team.