How quicksortrx/universal-med-ids works
Main Components
a. Data Sources Layer:
• OpenFDA Data Source
• RxNorm Database (rxnorm.db)
• FDA NDC Directory (package.csv, product.csv)
• Medicare Part B Data (optional component)
b. Core Processing Layer:
• QUMI Code Generator
• Data Transformation Pipeline
• Algorithmic Matching Engine
c. Output Layer:
• Universal Medication IDs (CSV format)
• Logging System
[Data Sources Layer]
• ↓
[Core Processing Layer]
• ↓
[Output Layer]
Detailed Component Relationships
a. Data Sources Layer:
[OpenFDA API] → [RxNorm DB] → [FDA NDC Files] → [Medicare Part B Files]
• ↓ ↓ ↓
• [Data Ingestion & Validation Layer]
b. Core Processing Layer:
[Data Transformation Pipeline]
• ↓
[Algorithmic Matching Engine]
• ↓
[QUMI Code Generator]
c. Output Layer:
[Universal Med IDs CSV] ← [Logging System]
Technical Details to Highlight
• Python-based implementation
• File-based data storage
• CSV output format
• Modular design with separate components for:
• Core QUMI generation
• Medicare pricing
• Logging
• Common utilities
Key Interactions
• Data flow from source files to QUMI generator
• Validation and transformation steps
• Error handling paths
• Logging points
• Optional Medicare pricing integration
Additional Elements
• Data flow directions
• Component types
• Validation points
• Optional vs. required components
• Error handling paths
This system design should clearly communicate both the data processing pipeline and the business rules that ensure accurate medication matching, which is crucial for the project's primary goal of providing universal medication identifiers.