Function testing

[victornyabote]

Add Function Testing for Time-Series Data Pipeline

🔑 Key Additions

• Implemented automated testing functions to validate the data ingestion and cleaning pipeline.
• Added rolling correlation analysis with a 20-period window across sensor datasets.
• Visualised raw vs. cleaned data to confirm correctness of preprocessing.
• Produced comparative plots showing correlation fluctuations and anomalies across datasets.

📊 Benefits

• Ensures the time-series pipeline is reliable and works consistently across multiple datasets.
• Provides validation evidence that cleaning improves stability and interpretability of sensor data.
• Detects anomalies (e.g., perfect correlation = 1.0) and supports further investigation.
• Builds foundation for backend integration and machine learning tasks in later stages.

🧪 Testing

• Verified that all four datasets (1321079.csv, 1350261.csv, 3036461.csv, 518150.csv) load and process correctly.
• Checked that rolling correlations compute without errors.
• Plots for raw vs. cleaned data confirm proper interpolation and formatting.

✅ Checklist

• Function testing notebook added/updated (sprint_2 continuation.ipynb).
• Pipeline tested across multiple sensor datasets.
• Rolling correlation analysis implemented.
• Visual outputs (raw vs. cleaned data) validated.