Net-Vis

Net-Vis is a web application designed to provide interactive visualizations of performance metrics for machine learning models in physics simulations. This project was built using React, TypeScript, and Python.

Features

• Interactive Visualizations: View and analyze performance metrics with dynamic charts and graphs.
• Data Processing CLI: Process and normalize data with a command-line interface.
• NetCDF File Support: Visualize fluid dynamics training data using matplotlib frame animations.
• Flask API Integration: Optional Flask server for managing and processing training metrics.

Screenshots

Installation

Prerequisites

• Node.js (v14 or later) for the client application.
• Python (v3.8 or later) for the server and data processing scripts.

Setup

1. Clone the Repository

   git clone https://github.com/josephjquinn/net-vis.git

2. Install Client Dependencies

   Navigate to the client directory and install the required Node.js packages:

   cd net-vis/client
   npm install

3. Install Server Dependencies

   Navigate to the server directory and install the necessary Python packages:

   cd ../server
   pip install -r requirements.txt

4. Run the Client Application

   Start the client application:

   npm run dev

5. Run the Data Processing Scripts

   Process your data using the CLI:

   python main.py [flags]

6. Start the Optional Flask Server

   If you wish to use the Flask server for API management:

   python run.sh

CLI Flags

The Data Processing Command Line Interface (CLI) provides several options for managing your data:

• --data_dir: Path to the directory containing data files (default: ./metrics/).
• --case_name: Specify a case to process; if omitted, all cases will be processed.
• --all: Process all cases in the directory.
• --norm: Normalize metrics for all cases.
• --min: Compute minimum values for all metrics across cases.

API Endpoints

If using the optional Flask server, the following API endpoints are available:

• GET /metrics/<case>/: Retrieve and process metrics for a specific case.
• GET /all/<case>/: Aggregate metrics from all cases in the directory.
• GET /all: Normalize metrics to a scale of 0-100 for radar charts.
• GET /norm: Extract minimum values for metrics across all cases.