SymPy is a Python library for symbolic mathematics: instead of only producing a number for one set of inputs, it can manipulate equations into formulas that remain useful for many inputs. It acts as a computer algebra system, letting code carry out algebra that would otherwise be tedious and error-prone by hand. In applied geometry, that makes mathematical notation executable without requiring advanced mathematics. section 1.5 section 1.7
The basic workflow is direct: create symbolic variables, state equations as expressions that equal zero, and pass both the equations and unknowns to `solve`. For two trains covering 450 km in one hour, with one traveling twice as fast as the other, SymPy solves the pair of constraints and returns a mapping from each symbol to its value. section 1.6
Listing 1.1 Numeric solution in SymPy
from sympy import *
Va, Vp = symbols('Va Vp') #1
solution = solve([
Vp - Va * 2,
Va * 1 + Vp * 1 - 450
], (Va, Vp))
print(solution)
This complete listing defines two symbolic speeds, expresses the distance and speed-ratio constraints, and solves them in one call. section 1.6
Replacing fixed values with symbols turns the same calculation into a reusable formula. A distance `D`, travel time `t`, and speed multiplier can remain symbolic, so the result describes both train speeds for any compatible inputs. SymPy can also render a result as Python code or LaTex, connecting derivation, implementation, and publication. section 1.6 section 1.7
For small linear-equation problems, symbolic solving offers practical benefits and limits:
- Use it to derive a general solution once, then use the generated formula instead of repeatedly calling a generic solver. section 3.7.3
- Prefer it for systems up to roughly 4 × 4; symbolic expressions grow in time and length factorially as dimensions increase. section 3.7.3
- Review generated code before adopting it, because it may need grooming for readability and performance. section 3.7.3 section 3.8.4
Least-squares polynomial coefficients
SymPy can solve a matrix equation for unknown polynomial coefficients even when the desired result is numeric. For rowing-session times, form the least-squares matrix and right-hand vector, represent the coefficients with symbols, solve the equation, then collect the returned dictionary values into a plain coefficient list for a general polynomial formula. This symbolic setup adds overhead compared with a numeric solver, but gives a uniform equation-solving workflow. section 6.2.3
