Best of Data Visualization — December 2024

  1. 1
    Article
    Avatar of communityCommunity Picks·1y

    Lightweight Charts™ library — TradingView

    TradingView's Lightweight Charts library offers a compact, feature-rich solution for interactive, high-performance charts. At just 45 KB, it uses HTML5 Canvas technology and supports streaming updates for custom data. The library is open-source, developer-friendly, and accessible, and is trusted by millions of traders and companies. It provides an easy-to-integrate charting solution with flexible styling options and is designed to be responsive and mobile-friendly.

  2. 2
    Article
    Avatar of communityCommunity Picks·1y

    One Million Screenshots

    Explore over a million rendered homepages from the web in an interactive manner, allowing you to zoom, pan, and click similar to Google Maps. This visual dataset could help you find websites you've been looking for or discover new ones. Check out the FAQ for more details and learn about the Screenshot API if you're interested in the data.

  3. 3
    Article
    Avatar of mlmMachine Learning Mastery·1y

    5 Tools for Visualizing Machine Learning Models

    Machine learning models require specialized tools to visualize their structure, performance, and behavior. Five useful tools for this purpose include TensorBoard for neural network models, SHAP for model prediction explanations, Yellowbrick for Python-based model diagnostics, Netron for deep learning model architecture visualization, and LIME for intuitive model explanations. These tools cater to various model types and use cases, helping users understand complex ML models better.

  4. 4
    Article
    Avatar of hnHacker News·1y

    anvaka/map-of-github: Inspirational Mapping

    This project maps over 400,000 GitHub projects, clustering them based on common stargazers. Using data from GitHub's public activity events, Jaccard Similarity, and AWS for processing, the map visualizes connections between repositories. The layout was computed with ngraph.forcelayout and rendered with maplibre. Clusters and labels were generated with the help of various tools, including ChatGPT. The project is open-source under the MIT license, with contributions welcome.

  5. 5
    Article
    Avatar of hnHacker News·1y

    Olshansk/postgres_for_everything: How to reduce complexity and move faster? Just Postgres for everything.

    Using PostgreSQL simplifies development by handling a wide variety of tasks including cron jobs, message queues, GIS & mapping, search, caching, and more. It shows PostgreSQL’s versatility and provides a repository of resources to use PostgreSQL for various purposes. Contributors are encouraged to submit new examples to enrich this collection.

  6. 6
    Article
    Avatar of mlmMachine Learning Mastery·1y

    How to Quickly Deploy Machine Learning Models with Streamlit

    Learn how to deploy a simple machine learning model using Streamlit to predict house prices based on size. The guide walks through generating synthetic data, training a linear regression model, and creating a user interface with Streamlit. It also covers how to deploy the model on Streamlit Cloud, making it accessible via a web app.

  7. 7
    Article
    Avatar of planetpythonPlanet Python·1y

    JupyterLab 101 Book is Now Available

    JupyterLab 101 is a comprehensive guide designed to familiarize users with JupyterLab's new user interface and functionalities. JupyterLab provides a tabbed interface for editing multiple Notebooks, opening terminals, creating a Python REPL, and more. It supports all the same programming languages as Jupyter Notebook and includes features like a debugger utility. The guide aims to help users efficiently navigate and utilize JupyterLab to produce high-quality Notebooks for teaching, presentations, data visualization, and machine learning.

See all Data Visualization archives