Omar Sobh
16161bb9df
The workspace root was upgraded to thiserror = "2" in an earlier commit, but 56 per-crate Cargo.toml files still independently declared "1.0". These crates do not use workspace.dependencies inheritance for thiserror. All updated to thiserror = "2" for complete fleet alignment. Includes: rtx-backend, rtx-tensor, rtx-losses, rtx-backend-cuda/rocm/metal, all training crates (rtx-auto, rtx-rl, rtx-distributed, rtx-federated, etc.), specialized crates (rtx-science, rtx-platform, rtx-nmf, rtx-neuro-*), production crates (rtx-streaming, rtx-serving-api), and all demo crates. cargo check --workspace: PASSES.
RustyTorch++ Demo UI
React + TypeScript + Tauri frontend for the RustyTorch++ demo platform. Features 18 interactive demos with GPU-accelerated backends.
Stack
- React 19 - UI framework
- TypeScript 5.6 - Type safety
- Three.js - WebGL 3D/2D visualization (via @react-three/fiber)
- Recharts - Data visualization and charts
- Tauri 2.x - Desktop app framework
- Vite 6 - Build tooling
- Vitest - Testing framework
Development
# Install dependencies
pnpm install
# Development server (React only)
pnpm dev
# Development with Tauri backend
pnpm tauri dev
# Production build
pnpm tauri build
Testing
# Run all tests
pnpm test
# Watch mode
pnpm test:watch
# Coverage report
pnpm test:coverage
Project Structure
src/
├── App.tsx # Main router component
├── main.tsx # React entry point
├── index.css # Global styles
│
├── pages/
│ ├── Home.tsx # Demo gallery landing page
│ └── demos/
│ ├── HemodynamicsDemo.tsx # Virtual Catheter
│ ├── MREDemo.tsx # MRE Elastography
│ ├── BioheatDemo.tsx # Thermal Ablation
│ ├── SlideScopeDemo.tsx # SlideScope Pathology
│ ├── NeuroDemo.tsx # RustyNeuro MEG/EEG
│ ├── NeuralOperatorDemo.tsx # Neural Operator PDE
│ ├── PIDDMDemo.tsx # Physics-Informed Diffusion
│ ├── DigitalTwinDemo.tsx # Medical Digital Twin
│ ├── FNOBenchmarkDemo.tsx # FNO Benchmark
│ ├── PINNBenchmarkDemo.tsx # PINN Benchmark
│ ├── InferenceProfilerDemo.tsx # Inference Profiler
│ ├── ModelZooDemo.tsx # Model Zoo
│ ├── PortfolioDemo.tsx # Portfolio Optimizer
│ ├── RiskAnalyzerDemo.tsx # Risk Analyzer
│ ├── TimeSeriesDemo.tsx # Time Series Forecast
│ ├── ImageClassifierDemo.tsx # Image Classifier
│ ├── ObjectDetectorDemo.tsx # Object Detector
│ └── SegmentationDemo.tsx # Segmentation
│
├── components/
│ ├── index.ts # Barrel exports
│ ├── DemoCard.tsx # Demo gallery card
│ ├── DemoLayout.tsx # Common demo page layout
│ │
│ ├── hemodynamics/ # Virtual Catheter components
│ │ ├── VesselCanvas.tsx # Three.js 2D vessel renderer
│ │ ├── ControlPanel.tsx # Simulation parameters
│ │ ├── Colorbar.tsx # Dynamic color legend
│ │ ├── PerformanceMonitor.tsx # FPS/inference metrics
│ │ ├── GeometryTools.tsx # Tool selection toolbar
│ │ ├── GeometryEditor.tsx # Interactive canvas overlay
│ │ ├── PulsatileControls.tsx # Cardiac cycle controls
│ │ └── ExportControls.tsx # Screenshot/data export
│ │
│ ├── mre/ # MRE Elastography components
│ │ ├── MRECanvas.tsx # Wave field visualization
│ │ ├── MREControls.tsx # Frequency/wavelength controls
│ │ └── StiffnessMap.tsx # Stiffness colormap
│ │
│ ├── bioheat/ # Thermal Ablation components
│ │ ├── TemperatureCanvas.tsx # 3D temperature field
│ │ ├── HeatSourcePanel.tsx # Source configuration
│ │ └── ThermalDoseDisplay.tsx # CEM43 visualization
│ │
│ ├── slidescope/ # SlideScope components
│ │ ├── SlideViewer.tsx # Deep zoom tile viewer
│ │ ├── StainChannels.tsx # Individual stain display
│ │ └── NMFProgress.tsx # Processing progress
│ │
│ ├── neuro/ # RustyNeuro components
│ │ ├── SignalViewer.tsx # EEG/MEG signal display
│ │ ├── TopoMap.tsx # Topographic mapping
│ │ └── ConnectivityMatrix.tsx # Connectivity visualization
│ │
│ ├── neural-operator/ # Neural Operator components
│ │ ├── PDECanvas.tsx # Interactive PDE canvas
│ │ ├── PDEControls.tsx # PDE type selection
│ │ ├── TrainingProgress.tsx # Training status
│ │ └── LossChart.tsx # Loss curve visualization
│ │
│ ├── piddm/ # PIDDM components
│ │ ├── DiffusionCanvas.tsx # Diffusion process viz
│ │ ├── SamplingControls.tsx # Sampling configuration
│ │ └── PhysicsLossChart.tsx # Physics loss tracking
│ │
│ ├── digital-twin/ # Digital Twin components
│ │ ├── OrganViewer.tsx # 3D organ visualization
│ │ ├── ProbePanel.tsx # Probe placement
│ │ └── SimulationResults.tsx # Simulation output
│ │
│ ├── benchmark/ # Benchmark components
│ │ ├── BenchmarkChart.tsx # Comparison charts
│ │ ├── MetricsTable.tsx # Results table
│ │ └── ConfigPanel.tsx # Benchmark configuration
│ │
│ ├── inference-profiler/ # Profiler components
│ │ ├── LatencyChart.tsx # Latency distribution
│ │ ├── ThroughputChart.tsx # Throughput over batch sizes
│ │ ├── MemoryChart.tsx # Memory usage
│ │ └── ProfilerControls.tsx # Model/device selection
│ │
│ ├── model-zoo/ # Model Zoo components
│ │ ├── ModelGrid.tsx # Model card grid
│ │ ├── ModelCard.tsx # Individual model display
│ │ └── InferencePanel.tsx # Run inference UI
│ │
│ ├── portfolio/ # Portfolio components
│ │ ├── EfficientFrontier.tsx # Frontier visualization
│ │ ├── AssetAllocation.tsx # Pie chart allocation
│ │ └── CorrelationMatrix.tsx # Asset correlations
│ │
│ ├── risk/ # Risk Analyzer components
│ │ ├── VaRChart.tsx # Value-at-Risk visualization
│ │ ├── StressScenarios.tsx # Stress test results
│ │ └── RiskDecomposition.tsx # Risk breakdown
│ │
│ ├── timeseries/ # Time Series components
│ │ ├── ForecastChart.tsx # Forecast visualization
│ │ ├── ModelSelector.tsx # Model type selection
│ │ └── MetricsDisplay.tsx # RMSE, MAE, MAPE
│ │
│ ├── vision/ # Vision components (shared)
│ │ ├── ImageUploader.tsx # Image upload/sample select
│ │ ├── ImageCanvas.tsx # Base image canvas
│ │ └── SampleImages.tsx # Sample image gallery
│ │
│ ├── image-classifier/ # Classifier components
│ │ ├── ClassificationResults.tsx # Top-K predictions
│ │ └── ConfidenceBar.tsx # Confidence visualization
│ │
│ ├── object-detector/ # Detector components
│ │ ├── DetectionCanvas.tsx # Bounding box overlay
│ │ ├── DetectionList.tsx # Detection results list
│ │ └── DetectorControls.tsx # Model/threshold controls
│ │
│ └── segmentation/ # Segmentation components
│ ├── MaskOverlay.tsx # Segmentation mask display
│ ├── ClassLegend.tsx # Class color legend
│ └── SegmentationControls.tsx # Model/class set controls
│
├── lib/
│ ├── types.ts # Common TypeScript types
│ ├── simulation.ts # Hemodynamics IPC wrappers
│ ├── colors.ts # Colormap utilities
│ ├── mre-types.ts # MRE types
│ ├── bioheat-types.ts # Bioheat types
│ ├── slidescope-types.ts # SlideScope types
│ ├── neuro-types.ts # RustyNeuro types
│ ├── neural-operator-types.ts # Neural Operator types
│ ├── piddm-types.ts # PIDDM types
│ ├── digital-twin-types.ts # Digital Twin types
│ ├── benchmark-types.ts # Benchmark types
│ ├── inference-profiler-types.ts # Profiler types
│ ├── model-zoo-types.ts # Model Zoo types
│ ├── portfolio-types.ts # Portfolio types
│ ├── risk-types.ts # Risk Analyzer types
│ ├── timeseries-types.ts # Time Series types
│ ├── image-classifier-types.ts # Image Classifier types
│ ├── object-detector-types.ts # Object Detector types
│ └── segmentation-types.ts # Segmentation types
│
└── data/
└── demos.ts # Demo registry with metadata
Demo Pages
Medical / Science
| Demo | Components | Features |
|---|---|---|
| Virtual Catheter | VesselCanvas, GeometryEditor, PulsatileControls | Three.js rendering, geometry editing, cardiac waveforms |
| MRE Elastography | MRECanvas, StiffnessMap | Wave field visualization, stiffness colormap |
| Thermal Ablation | TemperatureCanvas, HeatSourcePanel | 3D temperature field, thermal dose |
| SlideScope | SlideViewer, StainChannels | Deep zoom tiles, NMF stain separation |
| RustyNeuro | SignalViewer, TopoMap, ConnectivityMatrix | Multi-channel signals, source localization |
| Digital Twin | OrganViewer, ProbePanel | 3D organ mesh, interactive simulation |
AI / ML
| Demo | Components | Features |
|---|---|---|
| Neural Operator | PDECanvas, TrainingProgress, LossChart | Interactive boundary drawing, training UI |
| PIDDM | DiffusionCanvas, SamplingControls | Diffusion visualization, physics loss |
| FNO Benchmark | BenchmarkChart, MetricsTable | FNO vs FEM vs FDM comparison |
| PINN Benchmark | BenchmarkChart, ConfigPanel | Multi-PDE training comparison |
| Inference Profiler | LatencyChart, ThroughputChart, MemoryChart | Performance metrics, batch sweeps |
| Model Zoo | ModelGrid, InferencePanel | Model catalog, one-click inference |
Finance
| Demo | Components | Features |
|---|---|---|
| Portfolio Optimizer | EfficientFrontier, CorrelationMatrix | Markowitz frontier, asset allocation |
| Risk Analyzer | VaRChart, StressScenarios | VaR visualization, stress testing |
| Time Series | ForecastChart, ModelSelector | Multi-model forecasting |
Vision
| Demo | Components | Features |
|---|---|---|
| Image Classifier | ClassificationResults, ConfidenceBar | Top-K predictions |
| Object Detector | DetectionCanvas, DetectionList | Bounding boxes, NMS |
| Segmentation | MaskOverlay, ClassLegend | Pixel-wise masks, class colors |
IPC Communication
Tauri commands are invoked via @tauri-apps/api/core:
import { invoke } from "@tauri-apps/api/core";
// Example: Initialize profiler
const status = await invoke<ProfilerResponse>("profiler_initialize", { config });
// Example: Run detection
const result = await invoke<DetectionResult>("detect_objects", { imageData });
// Example: Start training
const handle = await invoke<string>("neural_operator_start_training", { config });
All IPC types are defined in src/lib/*-types.ts and mirror the Rust shared crate definitions.
Styling
Components use CSS-in-JS with inline styles for encapsulation. Global styles in index.css.
Color Scheme:
- Background:
#0f0f23(dark blue) - Panel background:
#16213e - Accent:
#3282b8(cyan) - Text:
#e4e4e4 - Success:
#4ade80 - Warning:
#f59e0b - Error:
#ef4444
Build
# Type check
pnpm type-check
# Lint
pnpm lint
# Build for production
pnpm build
Output: dist/ (web assets), src-tauri/target/release/ (desktop app)
Test Summary
| Category | Test Files | Tests |
|---|---|---|
| Hemodynamics | 9 | 290 |
| Neural Operator | 4 | 85 |
| Benchmark | 2 | 45 |
| Profiler | 3 | 65 |
| Vision | 6 | 120 |
| Finance | 4 | 80 |
| Total | 28 | 685 |
Known Issues
- Three.js chunk size warning (~600KB) - expected for WebGL library
- "Not implemented: navigation" in tests - JSDOM limitation for download links
- Canvas operations not fully testable in JSDOM - mock implementations used
Contributing
- Follow TDD: Write tests first
- Mirror Rust IPC types exactly in TypeScript
- Use existing component patterns
- Keep components focused and reusable