Supporting data/code for "Learning the One-Electron Reduced Density Matrix at SCF Convergence Thresholds"

This repository contains the datasets and analysis scripts used in our machine-learning (ML) based quantum chemistry study. The data and code are provided to ensure full reproducibility of the results reported in the associated manuscript.

Dataset:

• Training datasets: Includes training data for all molecules studied, stored in a structured and compact HDF5 format. 
• Testdata: Contains the complete test dataset used for evaluating the ML models.

Scripts/Code:

• RMSE analysis: Compute and analyze the root mean square error (RMSE) values for energies, dipoles, forces, and other properties.
• Learning curves:  Scripts and data for plotting the learning curves along training set sizes for some selected molecules.
• HOMO-LUMO Gap: Scripts and data for computing and plotting the HOMO-LUMO gaps.
• Force correction analysis: Scripts and data to compute and plot the force errors before and after correction.
• AIMD analysis: Scripts and data for analyzing ab initio molecular dynamics (AIMD) trajectories.
• Cost: Scripts and data for plotting both the training cost and AIMD cost.