logiStat: A Generalizable Simulation Analysis Tool

Operations research analysts across the national security enterprise rely on simulations that generate rich, multivariate datasets. However, the post-processing workflow to extract insights —requiring data wrangling, statistical modeling, and visualization—is often a time-consuming process that requires programming expertise. This presentation introduces logiStat, an open-source, interactive application designed to streamline and broaden access to simulation post-processing. Built with Python and Streamlit , logiStat is deployable on Posit Connect, the Department of War (DoW ) approved enterprise analytics platform, enabling secure and collaborative use. The tool is domain-agnostic; users upload any comma separated values (CSV) simulation output and, through an intuitive, no-code interface, can perform a full spectrum of analysis. Capabilities include exploratory data analysis via interactive correlation matrices and scatter plots, predictive modeling (Ordinary Least Squares, Polynomial, and Random Forest regression) with full diagnostic reporting, and statistical inference. To demonstrate its utility, the presentation will apply logiStat to output from a contested logistics simulation, showcasing how the tool rapidly identifies key drivers of sustainment outcomes. The goal of logiStat is to reduce the effort required to turn simulation output into usable analytical findings.