Alphabeta Math
TheoremStatement: AI-adaptedProof: AI-generated
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

If all partial derivatives exist on a neighbourhood and are continuous at a point, then the map is totally differentiable there with Jacobian derivative

Statement

Let U⊆Rm be open and let f:U→Rn. Suppose every partial derivative ∂jf exists on a neighbourhood of a∈U and is continuous at a. Then f is totally differentiable at a, and Df(a) is the linear map with matrix Jf(a).

Facts & Assumptions

Given: The stated neighbourhood existence and continuity hypotheses for all vector partial derivatives.

[L1]

Coordinate-by-coordinate increments stay inside a Euclidean ball and telescope the total increment (Small coordinate-by-coordinate increments stay inside a Euclidean ball and telescope the total increment).

[L2]

The vector mean-value inequality says ∥f(b)−f(a)∥2≤M(b−a) on a real interval when the derivative norm is bounded by M (The mean value inequality: if f:[a,b]→Rm is continuous and differentiable on (a,b) with ∥f′∥2≤M, then ∥f(b)−f(a)∥2≤M(b−a)).

[L4]

The space R0 is the zero space, its standard basis is empty, and its Euclidean norm is zero (The standard list e:n→Fn with ei(i)=1F and ei(j)=0F for j≠i is an ordered basis of Fn; hence dim⁡FFn=n, and F0 is the zero space with basis ∅ and dimension 0 clause 5, The Euclidean inner product ⟨x,y⟩=∑k<nxkyk on Rn). Coordinate matrices represent linear maps by their finite coordinate sums, including empty rows or columns.

[L5]

A Euclidean linear map preserves real linear combinations in all finite dimensions (A linear map L:Rm→Rn in Euclidean coordinates). Total differentiability is the normwise epsilon-delta remainder estimate, including its explicit zero-dimensional cases (The total (Fréchet) derivative Df(a) as the linear first-order approximation with o(∥h∥2) remainder).

Proof

technique · direct
1.1givenL4L5algebra

If m=0, then a=0, U={0}, and the only increment is h=0. The unique linear map L:R0→Rn has L0=0, so f(a+h)−f(a)−Lh=0 for every admissible increment. It satisfies the normwise remainder estimate with any δ>0. There are no coordinate directions; Jf(a) is the empty-column n×0 matrix of L. Thus the conclusion holds without dividing by m or invoking a punctured limit at an isolated point.

1.2givenL4L5algebra

If m≥1 and n=0, every value of f is the unique zero vector. The unique linear map L:Rm→R0 therefore leaves zero remainder for every increment, and its matrix, like Jf(a), has zero rows. This proves the conclusion directly, without applying the positive-codomain vector mean-value inequality. For the rest of the proof assume m,n≥1.

1.3givenL1L2

Choose a ball around a on which the partial derivatives exist. Given ε>0, continuity at a gives a smaller ball on which every ∥∂jf(z)−∂jf(a)∥2<ε/m.

2.1step 1.3L2algebra

For h in that smaller ball, [L1] writes the increment as coordinate segments. On each segment apply [L2] to the one-variable map obtained after subtracting the fixed linear term Jf(a); its derivative norm is at most ε/m.

3.1step 1.1step 1.2step 1.3step 2.1L3L5∎

Summing the segment bounds gives ∥f(a+h)−f(a)−Jf(a)h∥2≤(ε/m)∑j∣hj∣≤ε∥h∥2/m≤ε∥h∥2. Since ε is arbitrary, this verifies the normwise remainder condition and Jf(a) is Df(a). Together with the two zero-dimensional cases this proves the full Statement.

Depends on

Dependency tree · 2 levels

Sources