Tutorials

Welcome to the PERLA tutorials! These learning-oriented guides will help you understand how to work with the PERLA database and its tools.

Getting Started

New to PERLA? Start here to learn the basics:

• Getting Started - Step-by-step introduction to accessing and exploring the PERLA database in NOMAD

Jupyter Notebooks

These are a set of notebooks that performed analysis that we describe in the PERLA paper. They require a parquet file that you need to generate with the Query PERLA Database notebook.

• Query PERLA Database - Learn how to retrieve data programmatically via the NOMAD API
• Performance Evolution Analysis - Analyze power conversion efficiency trends over time
• Bandgap Evolution Analysis - Explore bandgap and composition trends in the literature
• Architecture Evolution Analysis - Track device architecture trends (n-i-p vs p-i-n structures)
• Diversity Analysis - Examine dataset composition and diversity metrics
• ML Prediction with CrabNet - Tutorial on machine learning predictions using perovskite data
• Distribution Shift Case Study - Analyze distribution shifts in ML models
• Paperbot Pipeline - Visualize the automated extraction pipeline
• Performance Check Filtering - Apply physics-based validation filters to data
• PERLA Evaluations Analysis - Analyze the performance of PERLA's data extraction evaluations

What's Next?

After completing the tutorials, check out the How-to Guides for task-specific instructions, or explore Explanation pages to understand the concepts behind PERLA.