/qm9_analysis

A simple code for analysis and modeling of QM9 dataset

Primary LanguagePython

QM9 Analysis

A sample code for analysis and modeling of QM9 dataset, including preprocessing adjustment based on qml and machine learning models.

  • Modeling molecular representation: Columb Matrix (CM), Bags of Bonds (BOB)
  • Machine learning: Kernel Ridge Regression, Random Forest.

Dataset

QM9 includes the spatial information and 12 quantum properties of 13385 small molecules for quantum chemistry research.

Usage

  • Download QM9 dataset

  • Install QML (Quantum machine learning) toolkit. Note that QML is built for analysis of QM7 (another dataset).

  • Adjust the data preprocessing process in QML. Replace the compound.py file in src of QML.

    mv qml/qml/compound.py compound_qm7.py
cp compound.py qml/qml/compound.py

  • Input the parameters in main.py. Choose Representation, Data size, Hyperparameter.

  • Run python main.py to perform validating and testing.

Code is generated by Github Copilot :) . What an amazing tool!