Quadratic polynomials (algebraic) refresher

A quadratic polynomial over a field \(F\) is a degree-2 polynomial

\[ f(x)=ax^2+bx+c,\qquad a\ne 0,\quad a,b,c\in F. \]

Over \(F=\mathbb{R}\) or \(\mathbb{C}\) this is the familiar object from school algebra, but many experiments treat \(f\) over other fields (like \(\mathbb{Q}\) or finite fields \(\mathbb{F}_p\)). [Dummit and Foote, 2004, Wikipedia contributors, 2026]

Completing the square

A standard normal form is obtained by completing the square:

\[ ax^2+bx+c = a\left(x+\frac{b}{2a}\right)^2 - \frac{b^2-4ac}{4a}. \]

This shows that (over a field of characteristic \(\ne 2\)) every quadratic is a shifted/scaled square plus a constant term.

Discriminant and factorization

The discriminant is

\[ \Delta=b^2-4ac. \]

Over a field where square roots make sense, the roots are

\[ x = \frac{-b \pm \sqrt{\Delta}}{2a}. \]

Algebraically, this means:

  • \(f\) factors over \(F\) iff \(\Delta\) is a square in \(F\) (when \(\mathrm{char}(F)\ne 2\)).

  • If \(\Delta\) is not a square, the polynomial is irreducible over \(F\).

Finite field case (\(\mathbb{F}_p\), odd prime \(p\))

Over \(\mathbb{F}_p\), “\(\Delta\) is a square” means \(\Delta\) is a quadratic residue mod \(p\). This is the bridge between algebra and number theory: quadratic residues, Legendre symbols, and Gauss sums.

Why quadratics show up in prime experiments

Quadratics are the simplest non-linear polynomials, so they are a natural testbed for “prime values of polynomials”:

  • Example: \(n^2+1\) (Landau’s 4th problem).

  • Example: \(n^2+n+41\) (Euler’s famous prime-producing streak).

Even when a polynomial produces many primes early on, modular obstructions (e.g., always divisible by some prime for some residue class) inevitably appear.

Practical numerical caveats

  • Stable quadratic formula: over floating point numbers, use a stable variant when \(b^2\gg 4ac\) to avoid catastrophic cancellation.

  • Exact arithmetic: for modular experiments, keep everything as integers mod \(p\); don’t mix floats.

  • Normalization: many plots look cleaner after shifting/scaling to the monic form \(x^2+Bx+C\) (when working over a field with \(a^{-1}\)).

References

See References.

[Dummit and Foote, 2004, Wikipedia contributors, 2026]