SeeNumPy - Flexible Array Visualizer

Overview

SeeNumPy is a Flask-based web application designed to offer flexible and intuitive visualisations for multi-dimensional NumPy arrays. This project focuses on providing users with a variety of tools and options to explore, customise, and analyse their data interactively. Whether you're working with simple 2D arrays or more complex 3D structures, SeeNumPy enables seamless exploration through a highly interactive UI.

Flask tutorial used: Web Development with Python Tutorial – Flask & Dynamic Database-Driven Web Apps

Features

• Upload & Validate .npz Files

  • Supports multiple arrays within a single .npz file.
  • Ensures all arrays are numeric and meet visualisation requirements.

• Dynamic & Flexible Visualisations

  • Array-Like Display: Intuitively explore arrays with structured expansion (downward, then rightward). (TODO)
  • Custom Views:
    • Switch between table views and transparent cube-like visualisations for 3D arrays. (TODO)
    • Enable heatmap views to represent data visually. (TODO)
  • User-Controlled Collapse/Aggregation: Choose specific levels to collapse and apply aggregations like sum, mean, or custom calculations dynamically. (TODO)

• Aggregations & Insights

  • Compute and display sum, mean, count, and other stats at different levels.
  • Aggregations adjust dynamically as users expand or collapse data.

• Reset & Start Fresh

  • A dedicated reset button clears the interface for a new session.

Installation & Setup

Requirements

Ensure you have Python 3.13 installed along with the required dependencies.

Steps

1. Clone the Repository

   git clone https://github.com/wayfinderza/see_numpy
   cd seenumpy
   

2. Create a Virtual Environment

   python -m venv .venv
   .venv\Scripts\Activate
   

3. Install Dependencies

   pip install -r requirements.txt
   
   

4. Run the Flask App

   python app.py
   

5. Navigate to the Web App
   • Open the web app.

Usage

1. Upload a .npz file containing one or more multi-dimensional arrays.
2. Select an array to explore from the list of uploaded arrays.
3. Use the expand/collapse controls to navigate through dimensions.
4. Choose from different view modes (e.g., heatmap, cube, table).
5. Apply aggregations or custom calculations at the desired levels.
6. Reset the session to begin a new exploration.

Roadmap & Issues

We are committed to expanding SeeNumPy's capabilities with more visualisation options and enhanced flexibility.

📌 Track Our Progress: GitHub Project Board