Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 2.2 (Picard–Lindelöf) — local existence, uniqueness, and existence time T₀ = min{T, δ/M}

Proved
TeschlODE.IVP.picard_lindelof

by mikedeng1 · Sep 28, 2026 · Mathlib 0df444a (Lean v4.33.1)

existenceordinary-differential-equationsp2o-batch-books5p2o-gran-per-chapterp2o-plan-bookp2o-v1picard-lindelofuniqueness

Let U⊆R×RnU \subseteq \mathbb{R} \times \mathbb{R}^nU⊆R×Rn be open, fff continuous on UUU, (t0,x0)∈U(t_0, x_0) \in U(t0​,x0​)∈U, and suppose fff is locally Lipschitz continuous in the second argument, uniformly with respect to the first, on UUU. Then:

  1. (local existence) there is ε>0\varepsilon > 0ε>0 and a solution xxx of x˙=f(t,x)\dot x = f(t,x)x˙=f(t,x) on (t0−ε,t0+ε)(t_0 - \varepsilon, t_0 + \varepsilon)(t0​−ε,t0​+ε) with graph in UUU and x(t0)=x0x(t_0) = x_0x(t0​)=x0​;
  2. (uniqueness) on every interval I∋t0I \ni t_0I∋t0​, any two solutions on III with graph in UUU and value x0x_0x0​ at t0t_0t0​ coincide on III;
  3. (existence time) if T,δ>0T, \delta > 0T,δ>0, V=[t0,t0+T]×Bˉδ(x0)⊆UV = [t_0, t_0 + T] \times \bar B_\delta(x_0) \subseteq UV=[t0​,t0​+T]×Bˉδ​(x0​)⊆U and M=max⁡V∣f∣M = \max_{V} |f|M=maxV​∣f∣, then there is a solution xxx with x(t0)=x0x(t_0) = x_0x(t0​)=x0​ on [t0,t0+T0][t_0, t_0 + T_0][t0​,t0​+T0​] whose graph stays in VVV (so ∣x(t)−x0∣≤δ|x(t) - x_0| \le \delta∣x(t)−x0​∣≤δ), where
T0=min⁡{T,δM},δM=∞ if M=0;(2.16)T_0 = \min\Bigl\{T, \frac{\delta}{M}\Bigr\}, \qquad \frac{\delta}{M} = \infty \text{ if } M = 0; \qquad (2.16)T0​=min{T,Mδ​},Mδ​=∞ if M=0;(2.16)
  1. the analogous statement holds on [t0−T0,t0][t_0 - T_0, t_0][t0​−T0​,t0​] with V=[t0−T,t0]×Bˉδ(x0)V = [t_0 - T, t_0] \times \bar B_\delta(x_0)V=[t0​−T,t0​]×Bˉδ​(x0​).

Together with uniqueness, item 3 says the unique solution exists at least on [t0,t0+T0][t_0, t_0 + T_0][t0​,t0​+T0​].

Formalization Note. The book writes V=[t0,t0+T]×Bδ(x0)V = [t_0, t_0+T] \times B_\delta(x_0)V=[t0​,t0​+T]×Bδ​(x0​) with the open ball and "the maximum of ∣f∣|f|∣f∣ on VVV", and says the solution "remains in Bδ(x0)B_\delta(x_0)Bδ​(x0​)". A maximum over the open ball need not exist, and the solution can reach the sphere ∣x−x0∣=δ|x - x_0| = \delta∣x−x0​∣=δ at t=t0+T0t = t_0 + T_0t=t0​+T0​ (e.g. n=1n = 1n=1, f≡1f \equiv 1f≡1, δ=1\delta = 1δ=1, T=2T = 2T=2). The statement therefore uses the closed ball Bˉδ(x0)\bar B_\delta(x_0)Bˉδ​(x0​), as the book's own proof on p. 37 does (C={x:∥x−x0∥≤δ}C = \{x : \|x - x_0\| \le \delta\}C={x:∥x−x0​∥≤δ}). MMM is given as the greatest element of {∣f(p)∣:p∈V}\{|f(p)| : p \in V\}{∣f(p)∣:p∈V}. T0T_0T0​ is written if M = 0 then T else min T (δ / M), the book's convention δ/0=∞\delta/0 = \inftyδ/0=∞. The book's final sentence says "[t0−T,t0][t_0 - T, t_0][t0​−T,t0​]"; the statement uses [t0−T0,t0][t_0 - T_0, t_0][t0​−T0​,t0​], the evident intended reading. "Some interval around t0t_0t0​" is read as ∃ε>0\exists \varepsilon > 0∃ε>0 with the open interval (t0−ε,t0+ε)(t_0 - \varepsilon, t_0 + \varepsilon)(t0​−ε,t0​+ε); "unique" is read as uniqueness on every interval containing t0t_0t0​ (sets III that are order-connected).

Preamble
import Mathlib
import Definitions.Def_TeschlODE_IVP_IsSolutionOn
import Definitions.Def_TeschlODE_IVP_LocallyLipschitzSecond
Formal statement
namespace TeschlODE.IVP

theorem picard_lindelof {n : ℕ} (U : Set (ℝ × EuclideanSpace ℝ (Fin n))) (hU : IsOpen U)
    (f : ℝ × EuclideanSpace ℝ (Fin n) → EuclideanSpace ℝ (Fin n)) (hf : ContinuousOn f U)
    (t₀ : ℝ) (x₀ : EuclideanSpace ℝ (Fin n)) (h₀ : (t₀, x₀) ∈ U)
    (hLip : LocallyLipschitzSecond U f) :
    -- local existence on an open interval around `t₀`
    (∃ ε : ℝ, 0 < ε ∧ ∃ x : ℝ → EuclideanSpace ℝ (Fin n), x t₀ = x₀ ∧
      IsSolutionOn U f (Set.Ioo (t₀ - ε) (t₀ + ε)) x) ∧
    -- uniqueness on every interval containing `t₀`
    (∀ I : Set ℝ, I.OrdConnected → t₀ ∈ I →
      ∀ x y : ℝ → EuclideanSpace ℝ (Fin n), x t₀ = x₀ → y t₀ = x₀ →
        IsSolutionOn U f I x → IsSolutionOn U f I y → Set.EqOn x y I) ∧
    -- existence time `T₀ = min {T, δ / M}` forward in time
    (∀ T δ M : ℝ, 0 < T → 0 < δ →
      Set.Icc t₀ (t₀ + T) ×ˢ Metric.closedBall x₀ δ ⊆ U →
      IsGreatest ((fun p => ‖f p‖) '' (Set.Icc t₀ (t₀ + T) ×ˢ Metric.closedBall x₀ δ)) M →
      ∃ x : ℝ → EuclideanSpace ℝ (Fin n), x t₀ = x₀ ∧
        IsSolutionOn (Set.Icc t₀ (t₀ + T) ×ˢ Metric.closedBall x₀ δ) f
          (Set.Icc t₀ (t₀ + (if M = 0 then T else min T (δ / M)))) x) ∧
    -- and backward in time
    (∀ T δ M : ℝ, 0 < T → 0 < δ →
      Set.Icc (t₀ - T) t₀ ×ˢ Metric.closedBall x₀ δ ⊆ U →
      IsGreatest ((fun p => ‖f p‖) '' (Set.Icc (t₀ - T) t₀ ×ˢ Metric.closedBall x₀ δ)) M →
      ∃ x : ℝ → EuclideanSpace ℝ (Fin n), x t₀ = x₀ ∧
        IsSolutionOn (Set.Icc (t₀ - T) t₀ ×ˢ Metric.closedBall x₀ δ) f
          (Set.Icc (t₀ - (if M = 0 then T else min T (δ / M))) t₀) x) := by sorry

end TeschlODE.IVP
Source
Teschl, Ordinary Differential Equations and Dynamical Systems (author's preliminary version of AMS GSM 140, 2012), p. 38, Theorem 2.2
Read-back

What the Lean code literally says, in plain math · claude-opus-5-5

Setting. Let n∈Nn \in \mathbb{N}n∈N and let Rn\mathbb{R}^nRn carry the Euclidean norm. The hypotheses are:

  • U⊆R×RnU \subseteq \mathbb{R} \times \mathbb{R}^nU⊆R×Rn is open;
  • f:R×Rn→Rnf : \mathbb{R} \times \mathbb{R}^n \to \mathbb{R}^nf:R×Rn→Rn is continuous on UUU;
  • (t0,x0)∈U(t_0, x_0) \in U(t0​,x0​)∈U;
  • fff satisfies the local Lipschitz condition on UUU in the following sense. For every compact V0⊆UV_0 \subseteq UV0​⊆U there is a real LLL with ∥f(t,x)−f(t,y)∥≤L∥x−y∥\|f(t, x) - f(t, y)\| \le L \|x - y\|∥f(t,x)−f(t,y)∥≤L∥x−y∥ whenever (t,x),(t,y)∈V0(t, x), (t, y) \in V_0(t,x),(t,y)∈V0​.

Solution notion. Below, "xxx solves the equation on JJJ with graph in WWW" means the following, for x:R→Rnx : \mathbb{R} \to \mathbb{R}^nx:R→Rn:

  • for every t∈Jt \in Jt∈J, (t,x(t))∈W(t, x(t)) \in W(t,x(t))∈W;
  • for every t∈Jt \in Jt∈J, xxx has derivative f(t,x(t))f(t, x(t))f(t,x(t)) at ttt relative to JJJ. At an endpoint of JJJ that belongs to JJJ, this derivative is one-sided. At an isolated point of JJJ it holds automatically.

Conclusion. The theorem asserts all four of the following.

  1. Local existence. There are ε>0\varepsilon > 0ε>0 and x:R→Rnx : \mathbb{R} \to \mathbb{R}^nx:R→Rn such that x(t0)=x0x(t_0) = x_0x(t0​)=x0​ and xxx solves the equation on the open interval (t0−ε,t0+ε)(t_0 - \varepsilon, t_0 + \varepsilon)(t0​−ε,t0​+ε) with graph in UUU.

  2. Uniqueness. Let I⊆RI \subseteq \mathbb{R}I⊆R be any order-connected set containing t0t_0t0​: an interval of any type, open, closed, half-open or unbounded, including the single point {t0}\{t_0\}{t0​}. Let x,y:R→Rnx, y : \mathbb{R} \to \mathbb{R}^nx,y:R→Rn satisfy x(t0)=y(t0)=x0x(t_0) = y(t_0) = x_0x(t0​)=y(t0​)=x0​, and suppose both solve the equation on III with graph in UUU. Then x(t)=y(t)x(t) = y(t)x(t)=y(t) for all t∈It \in It∈I.

  3. Forward existence time. Let T,δ,MT, \delta, MT,δ,M be reals with T>0T > 0T>0 and δ>0\delta > 0δ>0. Write Bˉ(x0,δ)\bar B(x_0, \delta)Bˉ(x0​,δ) for the closed Euclidean ball of radius δ\deltaδ around x0x_0x0​, and define the box

B+=[t0,t0+T]×Bˉ(x0,δ).B^+ = [t_0, t_0 + T] \times \bar B(x_0, \delta).B+=[t0​,t0​+T]×Bˉ(x0​,δ).

Assume B+⊆UB^+ \subseteq UB+⊆U, and assume MMM is the maximum of ∥f∥\|f\|∥f∥ on B+B^+B+: MMM is attained at some point of B+B^+B+ and bounds ∥f∥\|f\|∥f∥ on all of B+B^+B+. Set

T0={Tif M=0,min⁡{T, δ/M}if M≠0.T_0 = \begin{cases} T & \text{if } M = 0, \\ \min\{T,\ \delta/M\} & \text{if } M \ne 0. \end{cases}T0​={Tmin{T, δ/M}​if M=0,if M=0.​

Then there is x:R→Rnx : \mathbb{R} \to \mathbb{R}^nx:R→Rn with x(t0)=x0x(t_0) = x_0x(t0​)=x0​ that solves the equation on [t0,t0+T0][t_0, t_0 + T_0][t0​,t0​+T0​] with graph in the box B+B^+B+ (not merely in UUU).

  1. Backward existence time. The same statement holds with the box B−=[t0−T,t0]×Bˉ(x0,δ)B^- = [t_0 - T, t_0] \times \bar B(x_0, \delta)B−=[t0​−T,t0​]×Bˉ(x0​,δ) in place of B+B^+B+. Assume B−⊆UB^- \subseteq UB−⊆U and M=max⁡B−∥f∥M = \max_{B^-} \|f\|M=maxB−​∥f∥, and define T0T_0T0​ by the same formula. Then there is xxx with x(t0)=x0x(t_0) = x_0x(t0​)=x0​ that solves the equation on [t0−T0,t0][t_0 - T_0, t_0][t0​−T0​,t0​] with graph in B−B^-B−.

Degenerate cases.

  • Value of MMM: in parts 3 and 4, MMM is a value of a norm, so M≥0M \ge 0M≥0. The case M=0M = 0M=0 is handled separately, so no division δ/0\delta/0δ/0 occurs.
  • Uniqueness on a single point: with I={t0}I = \{t_0\}I={t0​}, uniqueness is immediate from x(t0)=y(t0)x(t_0) = y(t_0)x(t0​)=y(t0​). The empty III is excluded because t0∈It_0 \in It0​∈I is assumed.
  • Values outside the interval: the solutions produced in parts 1, 3 and 4 are arbitrary outside their intervals.
  • n=0n = 0n=0: R0\mathbb{R}^0R0 is one point, so every derivative condition holds automatically. The statement then concerns only membership of the graph in UUU, B+B^+B+ or B−B^-B−.
Human review
  • Endorsed by Shuze Chen · Oct 2, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 2, 2026

    Confirmed by the mission captain (proposal self-audit).

  • Endorsed by WillR · Nov 15, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me