How amriteren-debug/ocean-embed works
Ocean Predictor is a web application for exploring date-specific ocean temperature predictions across depth, with map visualization and sector-focused analysis. The main workflow is a user selecting a date, the frontend requesting raw surface data and a subsurface prediction, and the backend returning depth grids for visualization and analysis. The repository also includes ARGO dashboard and skill-metric capabilities, regional proof-of-concept views, embeddings, and model training. The README is empty, and excerpts are partial; links are limited to documented endpoints and visible calls.
Main entry points: main.py Prediction API · App.jsx Dashboard app · VisualizationCanvas.jsx Map visualization · train_pipeline.py Training pipeline
Components 20
Ocean explorer
- Dashboard app
frontend/src/App.jsx - Ocean API client
frontend/src/api/oceanApi.js - Map visualization
frontend/src/components/VisualizationCanvas.jsx - Profile popup
frontend/src/components/ProfilePopup.jsx
Prediction service
- Prediction API
backend/main.py - Ocean preprocessing
backend/preprocessing.py - Prediction cache
backend/main.py - Inference runner
backend/inference_pipeline.py
Analysis views
- Sector analysis
frontend/src/hooks/useSectorAnalysis.js - Sector insights
frontend/src/components/SectorInsightsCard.jsx - ARGO dashboard
frontend/src/components/ArgoDashboard.jsx - Percentile products
backend/compute_percentiles.py - Regional demo
backend/poc_demo.py
Model development
- Training pipeline
backend/train_pipeline.py - Model definition
backend/verify_prediction.py - Data fetching
backend/data_fetcher.py - Depth statistics
backend/compute_target_stats.py
Other components
- Ocean user
- Copernicus data
- ARGO observations
Connections 19
- Ocean user → Dashboard app: selects date
- Dashboard app → Prediction API: requests depths
- Dashboard app → Prediction API: requests raw SST
- Dashboard app → Prediction API: requests prediction
- Dashboard app → Map visualization: renders grids
- Dashboard app → ARGO dashboard: opens view
- Map visualization → Sector analysis: uses metrics
- Map visualization → Profile popup: opens profile
- Sector insights → Sector analysis: uses analysis
- Prediction API → Ocean preprocessing: loads and regrids
- Prediction API → Prediction cache: reads and writes
- Prediction API → Copernicus data: downloads data
- Prediction API → ARGO observations: compares observations
- Training pipeline → Model definition: instantiates model
- Training pipeline → Data fetching: loads dataset
- Inference runner → Model definition: loads model
- Prediction API → Regional demo: serves regional demo
- Prediction API → Percentile products: serves percentiles
- Prediction API → Depth statistics: loads target stats