README.md

In wilsonfrantine/ggmodel: Plot model objects with ggplot

ggmodel – Visualize Base R Models with ggplot2 🚧

ggmodel is an R package aimed at transforming model outputs from base R into beautiful 2D visualizations using ggplot2. 📊 Currently, the package is in beta and includes one core function – ggsurface – designed to generate 2D surface plots from linear, generalized linear, and loess models.

✨ Features

• 🔍 2D Model Surface Plots – Visualize interactions between two predictors in your model.
• ⚙️ Compatible with Base R Models – Works with lm, glm, and loess objects.
• 🎨 Fully Customizable – Tailor colors, grid sizes, and legend labels using intuitive arguments.
• 🚀 Simple & Efficient – Minimal code to produce publication-ready surface plots.

📥 Installation

Install the development version of ggmodel from GitHub:

# Install remotes package if not already installed
install.packages("remotes")

# Install ggmodel (beta)
remotes::install_github("wilsonfrantine/ggmodel")

🚀 Quick Demo

Here’s what you can do with the current version of ggmodel:

# Load necessary libraries
library(ggmodel)

# Fit a basic linear model
model <- lm(mpg ~ wt + hp, data = mtcars)

# Visualize model surface
ggsurface(model, x.var = "wt", y.var = "hp", legend.title = "Miles per Gallon")

The plot will show how weight (wt) and horsepower (hp) interact to influence miles per gallon (mpg). 👉 For a full tutorial, visit the ggmodel project page.

🛠️ Function Details

ggsurface() – The Core of ggmodel

ggsurface is a versatile function that generates 2D model surface plots for regression models in R.

Supported Models: - lm – Linear Models - glm – Generalized Linear Models - loess – Local Polynomial Regression

Key Arguments: - m – Model object (lm, glm, or loess). - x.var – Name of the predictor variable for the x-axis. - y.var – Name of the predictor variable for the y-axis. - legend.title – Title for the legend (default: response variable name). - xgrid.size / ygrid.size – Grid resolution (default: 15). - low.col / high.col – Gradient colors for low and high surface values. - n.bins – Number of bins for discretizing the surface (default: 10). - scale.type – Prediction scale (link or response).

🔨 Roadmap

🔧 Although ggmodel currently focuses on surface plots, future updates will include: - 📊 Coefficient Plots – Visualize model coefficients with confidence intervals. - 📈 Residual Analysis – Expanded diagnostic plots for GLMs. - 🧩 Support for More Models – Compatibility with gam (Generalized Additive Models) and mixed models.

🤝 Contributing

Since ggmodel is in early development, contributions are encouraged and highly appreciated!

• 🐛 Found a bug? Open an issue!
• 🚀 Have ideas for new functions? Submit a pull request or request features.
• 💬 Feedback or suggestions? Let us know through GitHub Issues.

📜 License

This project is licensed under the GPL License.

🚧 Disclaimer

ggmodel is in beta. While functional, the package currently offers limited features. Expect new additions soon!

wilsonfrantine/ggmodel documentation built on Dec. 28, 2024, 6:48 a.m.