Skip to content

Diophantine Approximation

Diophantine approximation studies how closely real numbers can be approximated by rational numbers.

number-theory book chatgpt

Approximating Real Numbers by Rational Numbers

Diophantine approximation studies how closely real numbers can be approximated by rational numbers.

Given a real number α\alpha, one seeks rational fractions

pq \frac pq

such that

∣α−pq∣ \left| \alpha-\frac pq \right|

is very small.

The subject lies between number theory, analysis, and geometry. It investigates how arithmetic structure constrains approximation quality.

Rational Numbers and Density

The rational numbers are dense in the real line. Between any two distinct real numbers there exists a rational number.

Thus every real number can be approximated arbitrarily closely by rational numbers.

However, the important question is quantitative:

How small can the error become relative to the denominator qq?

For example,

π≈227 \pi\approx\frac{22}{7}

gives moderate accuracy, while

π≈355113 \pi\approx\frac{355}{113}

gives extraordinarily high accuracy.

The denominator sizes matter as much as the error itself.

Dirichlet Theorem

A foundational result is the following theorem.

Theorem (Dirichlet). For every irrational number α\alpha, there exist infinitely many rational numbers

pq \frac pq

such that

∣α−pq∣<1q2. \left| \alpha-\frac pq \right| < \frac1{q^2}.

∣α−pq∣<1q2 \left|\alpha-\frac{p}{q}\right|<\frac{1}{q^2}

This theorem guarantees unexpectedly strong rational approximations.

The proof uses the pigeonhole principle and is one of the classical applications of combinatorial reasoning in number theory.

Continued Fractions and Optimal Approximation

Continued fractions produce the best rational approximations.

If

pnqn \frac{p_n}{q_n}

is a convergent of the continued fraction expansion of α\alpha, then

∣α−pnqn∣<1qn2. \left| \alpha-\frac{p_n}{q_n} \right| < \frac1{q_n^2}.

Moreover, no fraction with smaller denominator approximates α\alpha more accurately.

Thus continued fractions encode the optimal approximation structure of irrational numbers.

Approximation of Quadratic Irrationals

Quadratic irrational numbers have periodic continued fractions.

For example,

2=[1;2‾]. \sqrt2=[1;\overline2].

Its convergents are

1,32,75,1712,… 1,\frac32,\frac75,\frac{17}{12},\dots

These approximations satisfy

∣2−pnqn∣≍1qn2. \left| \sqrt2-\frac{p_n}{q_n} \right| \asymp \frac1{q_n^2}.

Quadratic irrationals are badly approximable, meaning that rational approximations cannot improve substantially beyond the 1/q21/q^2 scale.

Liouville Numbers

Some numbers admit much better approximations.

A Liouville number is a real number α\alpha such that for every positive integer nn, there exist infinitely many rational numbers p/qp/q satisfying

∣α−pq∣<1qn. \left| \alpha-\frac pq \right| < \frac1{q^n}.

∣α−pq∣<1qn \left|\alpha-\frac{p}{q}\right|<\frac{1}{q^n}

Such numbers can be approximated extraordinarily closely by rationals.

The first explicit example was constructed by entity[“people”,“Joseph Liouville”,“French mathematician”]:

∑k=1∞10−k!. \sum_{k=1}^{\infty}10^{-k!}.

Liouville proved that every Liouville number is transcendental.

This was the first rigorous proof that transcendental numbers exist.

Roth Theorem

Liouville theorem was later improved dramatically.

Suppose α\alpha is an irrational algebraic number. Then approximations much better than

1/q2 1/q^2

cannot occur infinitely often.

The strongest form is Roth theorem.

Theorem (Roth). Let α\alpha be an irrational algebraic number. For every ε>0\varepsilon>0, the inequality

∣α−pq∣<1q2+ε \left| \alpha-\frac pq \right| < \frac1{q^{2+\varepsilon}}

has only finitely many rational solutions.

Thus algebraic irrational numbers cannot be approximated “too well.”

This theorem is one of the deepest results in Diophantine approximation.

Simultaneous Approximation

One may also approximate several real numbers simultaneously.

Given

α1,α2,…,αn, \alpha_1,\alpha_2,\dots,\alpha_n,

one seeks integers q,p1,…,pnq,p_1,\dots,p_n such that

∣qαi−pi∣ \left| q\alpha_i-p_i \right|

is small for all ii.

This leads to higher-dimensional lattice geometry and Minkowski theory.

Simultaneous approximation is central in modern geometry of numbers and homogeneous dynamics.

Geometry of Numbers

Diophantine approximation has a natural geometric interpretation.

The rational approximation

pq \frac pq

corresponds to the lattice point

(q,p). (q,p).

Approximating α\alpha means finding lattice points close to the line

y=αx. y=\alpha x.

Thus approximation problems become lattice problems.

This viewpoint was developed systematically by entity[“people”,“Hermann Minkowski”,“German mathematician”] and became the foundation of the geometry of numbers.

Metric Diophantine Approximation

Another branch studies approximation properties of “almost all” real numbers.

For example:

  • almost all real numbers satisfy Dirichlet-type bounds,
  • almost all numbers are not badly approximable,
  • almost all continued fraction coefficients are unbounded.

These questions involve probability, measure theory, and ergodic theory.

The resulting subject is called metric Diophantine approximation.

Modern Perspective

Diophantine approximation now interacts with many advanced areas:

  • transcendence theory,
  • ergodic theory,
  • homogeneous dynamics,
  • modular forms,
  • arithmetic geometry,
  • dynamical systems.

The subject begins with elementary questions about fractions and irrational numbers but ultimately leads to deep structural phenomena involving symmetry, geometry, and arithmetic complexity.