Time Series Analyzer

A web-based tool for analyzing and visualizing time series data.

Features

• Upload and process time series data in various formats (CSV, Excel, JSON)
• Interactive visualization of time series data
• Statistical analysis and anomaly detection
• Forecasting using various models (ARIMA, SARIMA, Exponential Smoothing, Prophet, LSTM)
• Export results and visualizations
• Robust error handling and data validation
• Comprehensive JSON serialization for complex data structures

Installation

1. Clone this repository
2. Create a virtual environment:

   python -m venv venv
   

3. Activate the virtual environment:
   • On Windows: venv\Scripts\activate
   • On macOS/Linux: source venv/bin/activate
4. Install dependencies:

   pip install -r requirements.txt
   

Usage

1. Activate the virtual environment
2. Run the application:

   python run.py
   

3. Open your browser and navigate to http://localhost:5001

API Endpoints

The application provides several API endpoints:

• /api/enhanced-eda - Perform enhanced exploratory data analysis
• /api/minimal-forecast - Generate time series forecasts with various models
• /api/debug-json - Test JSON serialization with various data types
• /upload - Upload time series data files

Project Structure

TimeSeriesAnalyzer/
├── app/                    # Application package
│   ├── __init__.py         # Initialize the app
│   ├── routes.py           # Define routes
│   ├── models.py           # Data models
│   ├── analysis.py         # Analysis functions
│   ├── error_handling.py   # Error handling utilities
│   └── dependency_check.py # Dependency validation
├── data/                   # Sample data and user uploads
├── static/                 # Static files (CSS, JS)
│   ├── css/                # CSS files
│   └── js/                 # JavaScript files
├── templates/              # HTML templates
├── tests/                  # Unit tests
│   ├── test_enhanced_eda.py       # Tests for EDA functionality
│   ├── test_debug_json.py         # Tests for JSON serialization
│   └── test_error_handling.py     # Tests for error handling
├── docs/                   # Documentation
│   ├── model_documentation.md     # Model documentation
│   └── error_handling.md          # Error handling guide
├── venv/                   # Virtual environment
├── .gitignore              # Git ignore file
├── requirements.txt        # Project dependencies
└── run.py                  # Application entry point

Key Components

JSON Serialization

The application includes a robust JSON serialization system that handles:

• NaN and Infinity values
• NumPy arrays and data types
• Pandas Series and DataFrames
• Nested data structures
• Custom objects

Error Handling

Comprehensive error handling includes:

• Structured error responses
• Detailed logging
• User-friendly error messages
• Dependency validation

Forecasting Models

The following forecasting models are available:

• ARIMA (Auto-Regressive Integrated Moving Average)
• SARIMA (Seasonal ARIMA)
• Exponential Smoothing
• Prophet (Facebook's time series forecasting tool)
• LSTM (Long Short-Term Memory neural networks)

License

MIT