Define storage-ready internal time-series data model

[BytecodeBrewer]

Goal

Define a storage-ready internal data model for historical exchange-rate and market time-series data.

Why

ARGUS will use multiple data sources such as Frankfurter, yfinance and possibly future APIs. These sources return different response formats.

The analytics, dashboard, reporting, prediction and database layers should not depend directly on external API response structures.

A clear internal model makes it easier to normalize data, calculate metrics, store historical data and later move from local storage to server or cloud infrastructure.

Scope

• Define the internal structure for time-series data
• Define required fields such as:
  • date / timestamp
  • source
  • asset type
  • base currency or symbol
  • target currency or quote asset
  • value / rate / close price
  • optional metadata
• Explain how external clients should normalize API responses
• Explain how the model maps to:
  • pandas DataFrames
  • database tables
  • future cloud storage
  • report generation
  • prediction input datasets
• Propose a minimal first database table structure
• Document trade-offs and open questions

Acceptance criteria

• Internal time-series format is documented
• Required fields are defined
• Client normalization rules are documented
• Mapping to pandas is explained
• Mapping to a first database table is proposed
• Future cloud/storage considerations are mentioned
• Relationship to reports and predictions is explained
• No production implementation is required unless small helper types are useful

Note

Priority: Must