Replication Materials for "Automatic Bayesian Mixture Modeling for Multimodal Categorical Data"

Replication Materials for "Automatic Bayesian Mixture Modeling..." This package contains all code and data necessary to reproduce the figures, tables, and results presented in the manuscript. Requirements:- R >= 4.0- R packages: see 0. SETUP in analysis.R File Structure:Replication/├── README.md # This file├── analysis.R # Main script to reproduce all results├── MultiModalR_1.0.0.tar.gz # From devtools::build()├── data/│ └── SMG_fafaj_oa.csv # Real-world ecological data├── output/ # Created by script└── mpa_output/ # Created by script Installation:The replication package includes the complete source code for MultiModalR (MultiModalR_1.0.0.tar.gz). To install it: install.packages("MultiModalR_1.0.0.tar.gz", repos = NULL, type = "source") How to Reproduce:1. Open R/RStudio and set your working directory to this folder: setwd("path/to/Replication") 2. Install the tar file: install.packages("MultiModalR_1.0.0.tar.gz", repos = NULL, type = "source") 3. Run the complete analysis: source("analysis.R") 4. All figures, tables, and session_info.txt are stored in output and mpa_output directories. Expected Runtime:Full benchmark (10 iterations): ~4 mins Notes on Replication: CSSA (Certainty-Separation-Stability Assessment):The CSSA analysis requires manual adjustment of prior_means for each of the nine categories (AA, BB, ..., II). The values used in the paper are pre-populated in the dff data frame within analysis.R. For full replication, the reviewer would need to run the MCMC for each category separately with the corresponding prior_means. MPA (Mode Plausibility Assessment):The MPA analysis runs 246 categories with 10,000 iterations each. Data Availability:The real-world ecological dataset (SMG_fafaj_oa.csv) is included in this replication package. The MultiModalR package source code is provided as MultiModalR_1.0.0.tar.gz. All data and code necessary to reproduce the figures and tables are contained within this archive.