SymPy is a bit of a niche so I don't know that many books based on it, specifically. In general, I don't know that many math/phys/engineering books that take the "computation first" approach. I'm sure they exist (e.g. using Maple or numerical using MATLAB). I guess you don't need to have a book that is explicitly designed for SymPy—you could follow any basic math/science textbook and reproduce complicated calculations and derivations using SymPy in parallel with the narrative of the book. I bet most of UGRAD physics will be one-liners...

Here are some links to the best computer-based-science stuff I was able to find in my bookmarks:

[ The official docs] tutorial https://docs.sympy.org/latest/tutorial/index.html + other docs https://docs.sympy.org/latest/index.html + SymPy links from the wiki https://github.com/sympy/sympy/wiki/External-SymPy-Media%2C-...

[ Reference of equations from advanced physics in a very condensed manner ] https://www.theoretical-physics.net/dev/index.html == https://www.theoretical-physics.net/dev/theoretical-physics.... It's mostly equations, but there are snippets of SymPy interspersed in there (written by Ondřej Čertík, who started SymPy)

And now for something in the other extreme—instead of exact equations, you can do physics with simulations:

[ Modelica ] Imagine you had an ODE (ordinary differential equations) solver with pre-defined "modules" for electrical, mechanical, thermal, and other systems. Modelica is not Python-based, but has its own language for describing variables, differential equations, and initial conditions. https://marcobonvini.com/modelica/2020/06/29/all-about-model... via https://news.ycombinator.com/item?id=23690788 Intro talk https://www.youtube.com/watch?v=-mvEUuc-sWE and a Modelica book https://mbe.modelica.university/