Least squares adjustment software

JuliaGrid is an easy-to-use power system simulation tool for researchers and educators provided as a Julia package.

Fast CPU and GPU Python implementations of Improved Kernel Partial Least Squares (PLS) by Dayal and MacGregor (1997) and Fast Partition-Based Cross-Validation With Centering and Scaling for XTX and XTY by Engstrøm and Jensen (2025). This package also includes options to use sample weights for PLS modeling.

MATGRID is an easy-to-use power system simulation tool for researchers and educators provided as a MATLAB package.

Computation of training set (X^T * W * X) and (X^T * W * Y) or (X^T * X) and (X^T * Y) in a cross-validation setting using the fast algorithms by Engstrøm and Jensen (2025).

Comparison of a few different methods for estimating surface normals in a point cloud.

Generalized least squares method which takes uncertainties in both dependent and independent variables.

Least squares adjustment software

Stata implementation of Weighted Fixed Effects (WFE) estimators for causal inference with panel data (Imai & Kim, 2021, Political Analysis).(Public Preview, testing...)(Disclaimer: CURRENTLY WIP)

Predicting Car prices given the various characteristics

Project conducted in STAT 4355.001.S22. Utilized the R Programming Language to determine a multi-linear model fitting to predict the number of bike rentals. Determined the appropriate attributes that significantly influenced the number of bike rentals. Collaboration with three other classmates.

Analysis and comparison of Weighted Least Squares (WLS) and Ordinary Least Squares (OLS). Covers heteroscedasticity, Feasible WLS, Huber robust regression, and validates methods with a Monte Carlo simulation

Software-defined GNSS positioning engine. Compute position entirely in software from raw RTCM3 MSM4 observations and broadcast ephemeris — Keplerian orbits, Saastamoinen tropo, Klobuchar iono, Hatch-smoothed pseudoranges, multi-constellation WLS. DGNSS and RTK planned.

An in-depth analysis of a gaming retailer's customer data to improve overall sales performance through enhanced understanding of customer behavior and loyalty patterns.

Models parcel prices across Pittsburgh to isolate the effect of location by controlling for parcel characteristics. Uses WLS regression and ANCOVA to estimate adjusted prices, then links price variation to crime, school quality, and geography.

ETA (estimated time of arrival) and MultiTimer, written in Vue.js

Various codes centered around problems in Bayesian inference, Bayesian Linear Regression, and Bayesian Non-Linear Parameter Estimation using the application of various Markov Chain Monte Carlo Algorithms