GPTPlot

AI-powered scientific plotting tool inspired by gnuplot

GPTPlot combines the simplicity of gnuplot's CLI with the power of Python's scientific stack (matplotlib, pandas, seaborn). Create publication-quality plots from the command line or interactive shell.

✨ Features

• 🚀 Fast CLI plotting - Generate plots without writing code
• 📊 Multiple plot types - Line, scatter, heatmap, quiver, surface, and more
• 🎨 Beautiful themes - Journal, presentation, and dark modes
• 📁 Smart data loading - Auto-detects CSV, DAT, TXT formats
• ⚙️ YAML configs - Reproducible plots via configuration files
• 🔢 Interactive mode - Explore data with an interactive shell
• 📈 Data analysis - Built-in correlation, fitting, and statistics
• 🤖 AI-Ready - Designed for future LLM integration

🔧 Installation

pip install gptplot

Requires Python 3.9+

🚀 Quick Start

Basic plotting

# Simple line plot
gptplot data.csv --x time --y voltage

# Auto-detect columns and plot type
gptplot data.csv

# Create scatter plot with custom styling
gptplot data.csv --type scatter --theme journal -o figure1

Multiple plot types

# Heatmap
gptplot grid.dat --x 1 --y 2 --z 3 --type heatmap --cmap viridis

# Histogram
gptplot data.csv --y values --type hist --bins 50

# Box plot
gptplot data.csv --x category --y measurement --type box

Configuration files

# Create plot_config.yaml
cat > plot_config.yaml << EOF
type: scatter
x: time
y: voltage
xlabel: Time (s)
ylabel: Voltage (V)
theme: journal
dpi: 300
EOF

# Use config
gptplot data.csv --config plot_config.yaml

Interactive mode

gptplot data.csv --interactive

# In the shell:
> plot x=time y=voltage type=line
> set theme dark
> save my_plot
> quit

📊 Examples

Data Science Workflow

# Correlation heatmap
gptplot data.csv --corr

# Statistical summary
gptplot data.csv --summary

# Polynomial fitting
gptplot data.csv --x time --y signal --fit poly2

Scientific Plotting

# Vector field (quiver plot)
gptplot spins.dat --no-header \
  --x 1 --y 2 --u 4 --v 5 --z 6 \
  --type quiver --cmap coolwarm

# 3D surface
gptplot grid.dat --x 1 --y 2 --z 3 \
  --type surface --cmap viridis

🎨 Themes

Built-in professional themes:

--theme journal        # Clean, publication-ready
--theme presentation   # Large fonts, high contrast
--theme dark           # Dark background
--theme notebook       # Jupyter-style

📁 Supported Formats

• Input: CSV, DAT, TXT (auto-detected delimiters)
• Output: PNG, PDF, SVG
• Config: YAML, JSON

🛠️ Advanced Features

Custom output directory

gptplot data.csv -o myplot --output-dir figures/

File naming schemes

# Overwrite (default)
gptplot data.csv -o plot

# Timestamp (experiment tracking)
gptplot data.csv --save-naming timestamp -o experiment

# Numbered (multiple runs)
gptplot data.csv --save-naming numbered -o run

Column specification

# By name
gptplot data.csv --x time --y voltage

# By 1-based index
gptplot data.dat --no-header --x 1 --y 2

🤖 Future: AI Integration

GPTPlot is designed with future LLM integration in mind:

# Coming soon!
gptplot --llm "plot voltage vs time with a dark theme"

📖 Documentation

Full documentation available at: GitHub Repository

🤝 Contributing

Contributions welcome! Please feel free to submit a Pull Request.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgements

Inspired by gnuplot's simplicity and powered by Python's scientific stack:

• matplotlib
• pandas
• seaborn
• numpy
• scipy

📬 Contact

• GitHub: @arnobmukherjee1988
• Email: arnobmukherjee1988@gmail.com

---

Made with ❤️ for scientists and data enthusiasts