FrankenRobots vs SymPy Editor

Study two demanding robot-control problems from inside the optimization loop. FrankenRobots focuses on a Unitree G1 humanoid and a household robot arm. Both experiences bundle their simulation assets and numerical engine with the app, so the lab works without an account or remote service. In the humanoid lab, inspect a high-dimensional policy, change terrain and disturbances, watch recovery behavior, and follow how CMA-ES searches a 5,040-parameter space. Curriculum controls expose how the task changes as balance, locomotion, and robustness are combined. In the arm lab, work with a compact 128-parameter policy across household manipulation tasks. Change the target and objective, inspect trajectories, and compare candidate behavior during the search. FrankenRobots is a simulation and education tool. It does not connect to, command, or make safety claims about physical robots. All simulation state stays on the device.

Edit mathematics the way you read it. SymPy editor draws your expression as a real formula, and every piece of it responds to touch: tap a term to select it, type over it, pull it apart, or apply any function from SymPy, the Python computer algebra system that powers it all. The whole CAS runs on your phone. No account, no ads, no tracking, no network: CPython and SymPy are packaged inside the app, everything you type stays on your device, and it all works in airplane mode. EDIT FORMULAS, NOT CODE • Tap any sub-expression (a term, an exponent, a matrix entry) and edit it in place • Type over a selection, or place a caret between terms and insert • Tap an operator to change it: turn + into a product, make an equation, or delete it • Unwrap (cos θ → θ), isolate, delete, copy and paste sub-expressions • The SymPy source line under the formula follows it, and the formula follows the line: edit whichever you prefer ALL OF SYMPY, ONE TAP AWAY • Transform menu: simplify, expand, factor, partial fractions, trig rewriting and more • A menu per type: matrices offer determinant, inverse and transpose; integrals, sums, derivatives and limits offer evaluation; equations swap sides or collect everything on the left • A search box over every SymPy function, asking for the parameters each one needs • An "unevaluated" switch builds Integral(f, x) or Determinant(M) symbolically instead of computing it • A Symbols panel for assumptions (positive, real, integer…) and matrix symbols with their shapes EVERY STEP IS REMEMBERED • The history shows each step and what changed: in red what went, in green what came • Play it as a slideshow, at the speed you choose • Save it as a self-contained web page that works offline, or as a Python script that rebuilds every step with SymPy • Keep several sessions, each with its own history, all on your device FOR STUDENTS, TEACHERS AND SCIENTISTS Complete the square in front of a class, evaluate limits and integrals, manipulate matrices and N-dimensional arrays, and export the derivation when it is done. FREE AND OPEN SOURCE BSD 3-Clause, no strings attached. Worked derivations, the source code and a browser version are at https://upabjojr.github.io/sympy-editor/ — and the same editor runs in Jupyter notebooks (pip install sympy-editor).