Figures4Papers
My Python scripts to make high-quality figures for publications in top AI conferences and journals.
Last verified:
What is Figures4Papers?
Figures4Papers is a collection of Python scripts created by Chen Liu, a Computer Science PhD Candidate at Yale University, for making high-quality, publication-ready scientific figures. The repository centralizes the author's own figure-making code that he developed for his research papers published in top AI conferences and journals like NeurIPS, ICML, ICASSP, and Nature Machine Intelligence.
The tool uses matplotlib to generate publication-quality figures with proper formatting for academic publications. It includes pre-configured scripts for various chart types commonly used in research papers, with styling that matches the requirements of top-tier AI conferences and scientific journals. The scripts handle figure sizing, font selection, color schemes, and other visual elements to ensure consistency and professionalism.
Key features include templates for multi-panel figures, support for scientific plotting conventions (log scales, proper ticks, colorbars), and reusable code that can be adapted with custom data. The repository has gained over 1,700 GitHub stars, indicating its usefulness to the research community.
This tool is primarily for researchers, PhD students, and data scientists who need to create publication-quality figures for academic papers, especially in AI/ML fields. It is particularly valuable for those who want to replicate the professional figure quality seen in top conference papers without starting from scratch.
Figures4Papers pricing
Pricing model: Freemium
Free - Open source under MIT License with no paid tiers
Figures4Papers pros
- Open source and completely free to use
- MIT License allows flexible usage
- 1,700+ GitHub stars prove community trust
- Created by Yale PhD candidate with top conference publications
- Includes ready-to-use matplotlib scripts
- Proper formatting for NeurIPS, ICML, ICASSP requirements
- Multi-panel figure support with GridSpec
- Scientific styling with proper fonts and sizes
- Built-in colorbar placement utilities
- Log scale axis support
- Reusable templates for common chart types
- No installation required - just copy scripts
- Works with standard Python data science stack
- Includes example scripts for quick adaptation
- Proven in actual published papers
Figures4Papers cons
- Not a packaged library - just raw scripts
- Requires Python and matplotlib knowledge
- No graphical user interface
- Scripts are author's personal code, not general-purpose
- Limited documentation beyond README
- No automated error handling
- May need customization for specific use cases
- No active maintenance or support team
Frequently asked questions about Figures4Papers
What is Figures4Papers?
Figures4Papers is a centralized GitHub repository of Python scripts for creating high-quality, publication-ready scientific figures. It contains the author Chen Liu's own figure-making code developed for his research papers published in top AI conferences like NeurIPS, ICML, and journals like Nature Machine Intelligence.
Is Figures4Papers free to use?
Yes, Figures4Papers is completely free and open source under the MIT License. Anyone can copy, modify, and use the scripts for their own research without any cost or restrictions.
What Python packages do I need?
The scripts primarily require matplotlib for plotting. You may also need numpy for data manipulation. These are standard Python data science packages available through pip or conda.
What types of figures are supported?
The repository includes scripts for various scientific figure types commonly used in research papers, including multi-panel figures using GridSpec, line plots, bar charts, scatter plots, figures with colorbars, and plots with log-scale axes.
Can I use these scripts for my own papers?
Yes, the MIT License allows you to freely use, modify, and distribute the scripts for any purpose, including your own research papers and publications.
Do the scripts work for non-AI research fields?
Yes, while the scripts were developed for AI/ML research papers, they are general-purpose matplotlib scripts that can be adapted for any scientific field requiring publication-quality figures.
How do I customize the scripts for my data?
You need to replace the data arrays, labels, colors, and file output paths in the scripts with your own data. The scripts are designed to be copied and adapted rather than used as a library.
What output formats are supported?
The scripts can save figures in multiple formats including PDF (recommended for publications), PNG, and other formats supported by matplotlib's savefig function.
Is there installation required?
No installation is needed. You simply clone or download the repository and copy the scripts you need into your project. The scripts run directly with Python and matplotlib.
Who created Figures4Papers?
Figures4Papers was created by Chen Liu, a Computer Science PhD candidate at Yale University working in the Krishnaswamy Lab. He is also the creator of CitationMap and a top reviewer at NeurIPS and ICML conferences.