hackingmaterials/matminer

Data mining for materials science

What it solves

Matminer is designed to simplify data mining in materials science, allowing researchers to easily access community-developed datasets and apply standardized methods for analyzing materials data.

How it works

The library provides a toolkit for retrieving data from various sources, applying featurizers to transform raw materials data into machine-learning-ready formats, and accessing a collection of community-curated datasets.

Who it’s for

It is primarily intended for materials scientists and researchers who want to use data mining and machine learning techniques to study materials properties.

Highlights

  • A comprehensive toolkit for materials data mining.
  • Integrated access to multiple community-developed datasets.
  • Built-in featurizers to prepare data for analysis.
  • Built-in citation tracking for datasets and featurizers to ensure original authors are credited.

Related

  • Project
    tensorflow/datasets

    A collection of ready-to-use public datasets provided as tf.data.Datasets to simplify the process of loading and preparing data for machine learning.

  • Project
    rasbt/mlxtend

    A Python library providing machine learning extensions for data science tasks, including ensemble methods, feature selection, and visualization utilities.

  • Project
    materialyzeai/maml

    A Python package that provides high-level interfaces for materials science machine learning, enabling the conversion of materials data into features for property prediction and energy surface modeling.

  • Project
    janosh/pymatviz

    A toolkit for materials informatics that provides specialized visualizations for chemical compositions, crystal structures, and phonon properties.

  • Project
    feature-engine/feature_engine

    A Python library providing Scikit-learn compatible transformers for comprehensive feature engineering and selection to improve machine learning model performance.