Description
Automated Statistical Analysis and Plotting with CLD.
Description
A lightweight tool that provides a reproducible workflow for selecting and executing appropriate statistical analysis in one-way or two-way experimental designs. The package automatically checks for data normality, conducts parametric (ANOVA) or non-parametric (Kruskal-Wallis) tests, performs post-hoc comparisons with Compact Letter Displays (CLD), and generates publication-ready boxplots, faceted plots, and heatmaps. It is designed for researchers seeking fast, automated statistical summaries and visualization. Based on established statistical methods including Shapiro and Wilk (1965) <doi:10.2307/2333709>, Kruskal and Wallis (1952) <doi:10.1080/01621459.1952.10483441>, Tukey (1949) <doi:10.2307/3001913>, Fisher (1925) <ISBN:0050021702>, and Wickham (2016) <ISBN:978-3-319-24277-4>.
README.md
statdecideR 
statdecideR is an R package for automatic statistical decision-making, compact letter displays (CLDs), and beautiful plots for parametric and non-parametric tests — making your analysis both robust and publication-ready.
✨ Features
- Automatically detects normality (Shapiro-Wilk test)
- Chooses appropriate test: ANOVA or Kruskal-Wallis
- Performs post-hoc test: Tukey HSD or Multiple Comparisons (agricolae)
- Displays Compact Letter Display (CLD) results
- Generates clean boxplots, heatmaps, and faceted plots
- Fully reproducible and customizable
📦 Installation
# Development version (from local source or GitHub in future)
# install.packages("devtools")
devtools::install_github("yourusername/statdecideR")
⚠ Replace
"yourusername"with your actual GitHub username once hosted.
🛠️ Usage
library(statdecideR)
# Load your data
data <- read.csv("your_data.csv")
# Run analysis and plot
plot_with_cld(
data = data,
dep_var = "pollen", # Your dependent variable
group_var = "month", # Your independent variable (factor)
normal = FALSE # Will be auto-detected in future version
)
📊 Example Output
The package automatically performs: - Statistical test selection - Post-hoc analysis (if significant) - Plot with CLD labels
📄 License
MIT © Subhradip Bhattacharjee.