// the take
A chapter-by-chapter working-through of Thrun/Burgard/Fox's 'Probabilistic Robotics' textbook, covering recursive state estimation through POMDPs (ch2-ch15) with math writeups as PDFs and code in C++, with some chapters also done in Python, MATLAB, and Julia. Useful for someone actually doing the exercises and wanting to check their derivation or implementation against a working one, especially for the harder later chapters (SEIF, FastSLAM) that most solution sets skip.
Goes deep into chapters most people give up on — SEIF and FastSLAM are correctly implemented and validated against the real Victoria Park dataset, not a toy simulation, with the SEIF output plotted to prove it converges. Several chapters have the same algorithm in two or three languages (C++/MATLAB/Julia/Python), which is genuinely handy for seeing how the same linear algebra reads in each. The PDFs pair the derivation with the code, so you can check your own math against the author's rather than just diffing code.
No top-level build — each chapter has its own makefile with no documented dependency versions for Armadillo, Intel MKL, or LAPACK, so getting even one chapter to compile means guessing. Binary dataset files (.h5, .bin) are committed straight into the repo instead of fetched or LFS-tracked, bloating every clone. Dead since May 2023, and the README links to a hardcoded GCP IP for a live demo that's almost certainly gone. No tests of any kind, so there's no way to tell if a given chapter's code still produces the plotted result without running it and eyeballing the output yourself.
We read the README, the directory tree and a key source file in October 2026. Nobody pays to be listed. How we judge →