We are excited to present this material, and all contributions are welcome. We will be using the materials project and it's available data to "predict" material properties through machine learning. Despite some examples being relatively simple, the main focus is on introducing the typical workflow in machine learning.
This set of notebooks is designed to familiarize you with the essential concepts of machine learning. Python has quickly become an excellent tool for these steps and we intend to show you the process with some methods from each branch of machine learning.
Python is a versatile language with many functionalities. It is a great glue language.
The packages that we will be using:
-
requests for gathering the materials project data
-
pandas for data management. A supercharged excel spreadsheet.
-
matplotlib for data visualization
-
numpy the foundation for pandas and linear algebra in Python
-
scikit-learn popular machine learning library, doesn't perform neural network calculations.
-
pymatgen a package by the Materials Project for working with material science structures and analysis of calculations.
Several additional resources we utilize are:
-
mybinder a tool for creating a custom programming environment hosted on google cloud.
-
materialsproject provides data from their VASP simulations available using a RESTful API.
To get started, launch the introduction notebook with binderhub.
All contributions, bug reports, bug fixes, documentation improvements, enhancements, and ideas are welcome! For these, please visit the Gitlab repository. Github is used only for better visibility.
Contributors:
- designhak (maintainer)
MIT