Design and Anomaly Detection in River Networks

Overview

This repository contains the code and data for the paper "Bayesian Design for Sampling Anomalous Spatio-Temporal Data" [1]. The primary objective of this project is to develop a robust Bayesian optimal experimental design (BOED) framework with anomaly detection methods to ensure high-quality data collection. The framework involves anomaly generation, detection, and error scoring within the search for optimal designs. See below for implementation for two simulated case studies: a spatial dataset and a spatio-temporal river network dataset.

Repository Structure

.
├── river_spatiotemporal_anomaly_design.R
├── spatial_anomaly_design.R
└── README.md

Getting Started

Prerequisites

Ensure you have the following software installed:

• R (version 3.0 or higher)
• RStudio (optional but recommended)
• Required R packages: oddstream, dplyr, dplyr, SSN, FNN

Installation

1. Clone this repository to your local machine:

   git clone https://github.com/KatieBuc/design_anom.git
   cd design_anom
           

2. Install the required R packages. Open R or RStudio and run:

   install.packages(c("oddstream", "dplyr", "dplyr", "SSN", "FNN"))

Usage

The scripts are designed to be run on a High-Performance Computing (HPC) cluster. Below are the steps to submit the scripts to the HPC.

River network example:

R -e "seed <- $seed; prop=$prop; lambda=$lambda; scenario=$scenario; source('./river_spatiotemporal_anomaly_design.R');

Contributing

If you would like to contribute to this project, please fork the repository and create a pull request with your changes. Contributions are welcome and appreciated!

License

This project is licensed under the MIT License. See the LICENSE file for more details.

Thank you for using this repository! We hope you find it useful for your research and projects.

¹ Buchhorn, Katie, et al. "Bayesian Design for Sampling Anomalous Spatio-Temporal Data." arXiv preprint arXiv:2403.10791 (2024). ↩