Integrate IPFS for Historical Data Storage and Retrieval

[PriyanshuVijay-S]

Overview

This pull request integrates IPFS to manage the storage and retrieval of historical data within our learning ABCI agents. By leveraging IPFS, our agents can now store data in a decentralized manner and fetch it as needed, enhancing data availability and integrity. This integration involves updates to behaviors, rounds, payloads, and introduces new functions for interacting with IPFS.

Changes

Detailed changes to the agent framework to incorporate IPFS are as follows:

Modified Files

• behaviours.py:

  • Added send_historical_data_to_ipfs in the EvolutionBehaviour class for data upload.
  • Added get_historical_data_from_ipfs in the TxPreparationBehaviour class for data retrieval.

• rounds.py:

  • Introduced historical_data_ipfshash and comparison_data attributes in the SynchronizedRound class to handle data synchronization and comparison based on the data fetched from IPFS.

• payload.py:

  • Added a new class EvolutionPayload to manage the payload structure for transactions involving IPFS interactions.

IPFS Integration

• Data Storage: Historical market data is now uploaded to IPFS via the send_historical_data_to_ipfs function, which returns an IPFS hash stored in the historical_data_ipfshash.
• Data Retrieval: The IPFS hash is used to fetch historical data using the get_historical_data_from_ipfs function during transaction preparations.

Impact

This integration:

• Enhances the robustness and decentralization of data storage.
• Improves the reliability and verifiability of historical data used in market analysis.
• Facilitates better synchronization and comparison of market data among agents, ensuring data integrity and consistency.