{
  "id": "exact-differential",
  "title": "Exact Differential",
  "lang": "en",
  "source_lang": "en",
  "summary": "An exact differential is a differential form that is the differential of a globally defined function, making its integral depend only on endpoints.",
  "category": "mathematics",
  "subcategory": "analysis",
  "type": "concept",
  "aliases": [
    "Total differential of a function"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/exact-differential/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/exact-differential/",
      "json": "https://aiwiki.page/zh/exact-differential.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "An exact differential is a differential form that can be written as the differential \\(df\\) of a single-valued function \\(f\\) on its entire domain. In several variables, an expression \\(\\omega=\\sum_i a_i\\,dx_i\\) is exact if there is a function \\(f\\) whose partial derivatives satisfy \\(a_i=\\partial f/\\partial x_i\\). Exactness connects differentiation with integration: integrating \\(df\\) along a path gives the difference between the values of \\(f\\) at the endpoints, independently of the path taken."
    },
    {
      "heading": "Definition and the total differential",
      "text": "Let \\(U\\) be an open subset of Euclidean space \\(\\mathbb{R}^n\\), and let \\(f:U\\to\\mathbb{R}\\) be continuously differentiable. Its total differential is\n\n\\[df=\\frac{\\partial f}{\\partial x_1}\\,dx_1+\\cdots+ \\frac{\\partial f}{\\partial x_n}\\,dx_n.\\]\n\nAt a point \\(p\\), the differential \\(df_p\\) is a linear functional on displacement vectors \\(h\\), given by\n\n\\[df_p(h)=\\sum_{i=1}^{n} \\frac{\\partial f}{\\partial x_i}(p)h_i.\\]\n\nIt describes the first-order change of \\(f\\):\n\n\\[f(p+h)-f(p)=df_p(h)+o(\\|h\\|).\\]\n\nThus the differential is not generally identical to a finite change in the function; it is the linear part of that change.\n\nConversely, suppose a differential one-form is given:\n\n\\[\\omega=\\sum_{i=1}^{n}a_i(x)\\,dx_i.\\]\n\nIt is exact on \\(U\\) if there exists a globally defined function \\(f\\) on \\(U\\) such that\n\n\\[\\omega=df.\\]\n\nThe function \\(f\\) is called a potential or primitive. On a connected open domain it is unique up to an additive constant. On a disconnected domain, the constant may be chosen separately on each connected component.\n\n“Total differential” and “exact differential” emphasize different questions. A total differential is constructed from a known function; exactness asks whether a given differential expression is the total differential of some function.\n\nIn one variable, this reduces to the familiar antiderivative problem. Every continuous expression \\(a(x)\\,dx\\) on an interval has a primitive. In several variables, compatibility conditions and the shape of the domain become essential."
    },
    {
      "heading": "Path independence and closed-loop integrals",
      "text": "Let \\(\\gamma:[a,b]\\to U\\) be a piecewise continuously differentiable curve. If \\(\\omega=df\\), the chain rule gives\n\n\\[\\int_\\gamma \\omega = \\int_a^b df_{\\gamma(t)}\\bigl(\\gamma'(t)\\bigr)\\,dt = f(\\gamma(b))-f(\\gamma(a)).\\]\n\nThis is the line-integral form of the fundamental theorem of calculus. It has two immediate consequences:\n\n- The line integral depends only on the endpoints, not on the path.\n- The integral around every closed curve is zero:\n\n\\[\\oint_\\gamma\\omega=0.\\]\n\nConversely, for a continuous one-form on a connected open subset of \\(\\mathbb{R}^n\\), path independence implies exactness. Fixing a base point \\(p_0\\), one defines\n\n\\[f(p)=\\int_{p_0}^{p}\\omega.\\]\n\nPath independence makes this definition unambiguous, and differentiation gives \\(df=\\omega\\). Equivalently, exactness can be characterized by the vanishing of integrals around all piecewise smooth closed curves.\n\nUsing the Euclidean inner product, a form\n\n\\[\\omega=F_1\\,dx_1+\\cdots+F_n\\,dx_n\\]\n\ncorresponds to a vector field \\(F=(F_1,\\ldots,F_n)\\). Exactness then means\n\n\\[F=\\nabla f,\\]\n\nwhere \\(\\nabla f\\) is the gradient of a scalar potential."
    },
    {
      "heading": "The mixed-partial test",
      "text": "In two variables, write\n\n\\[\\omega=M(x,y)\\,dx+N(x,y)\\,dy,\\]\n\nwith \\(M\\) and \\(N\\) continuously differentiable. If \\(\\omega=df\\), then\n\n\\[M=\\frac{\\partial f}{\\partial x}, \\qquad N=\\frac{\\partial f}{\\partial y}.\\]\n\nEquality of the mixed partial derivatives therefore requires\n\n\\[\\frac{\\partial M}{\\partial y} = \\frac{\\partial N}{\\partial x}.\\]\n\nIn \\(n\\) variables, the corresponding conditions are\n\n\\[\\frac{\\partial a_i}{\\partial x_j} = \\frac{\\partial a_j}{\\partial x_i} \\qquad\\text{for all }i,j.\\]\n\nUnder the stated regularity assumptions, these conditions are necessary for exactness. They are sufficient locally, but not necessarily on the whole domain.\n\nClosed versus exact forms\n\nThe exterior derivative expresses these compatibility conditions without choosing a particular coordinate system. In two variables,\n\n\\[d\\omega= \\left( \\frac{\\partial N}{\\partial x} - \\frac{\\partial M}{\\partial y} \\right)dx\\wedge dy.\\]\n\nA form is closed if \\(d\\omega=0\\). Since exterior differentiation satisfies\n\n\\[d(df)=0,\\]\n\nevery sufficiently regular exact form is closed.\n\nThe converse depends on the domain:\n\n- Every closed one-form is locally exact.\n- On a star-shaped open domain, every closed one-form is globally exact.\n- More generally, on a simply connected open domain, every closed one-form is exact.\n\nLocal exactness is an instance of the Poincaré lemma. A convex open domain is a particularly useful sufficient case, because it is star-shaped.\n\nThus the mixed-partial test must be used together with information about the domain. Equality of mixed partials alone does not establish global exactness."
    },
    {
      "heading": "A closed form that is not exact",
      "text": "On the punctured plane\n\n\\[U=\\mathbb{R}^2\\setminus\\{(0,0)\\},\\]\n\nconsider\n\n\\[\\omega= -\\frac{y}{x^2+y^2}\\,dx + \\frac{x}{x^2+y^2}\\,dy.\\]\n\nIts coefficients satisfy the mixed-partial conditions everywhere in \\(U\\), so it is closed.\n\nHowever, along the counterclockwise unit circle\n\n\\[x=\\cos t,\\qquad y=\\sin t,\\qquad 0\\leq t\\leq2\\pi,\\]\n\nthe pullback of \\(\\omega\\) is \\(dt\\). Consequently,\n\n\\[\\oint_\\gamma\\omega=2\\pi.\\]\n\nThe form cannot be exact on the entire punctured plane.\n\nLocally it is the differential of an angle coordinate, \\(\\omega=d\\theta\\). The obstruction is that no continuous, single-valued angle function exists on the whole punctured plane. Removing a suitable ray permits a branch of the angle and makes the restricted form exact.\n\nThis example shows how topology enters a differentiation problem: the local derivative conditions hold, but loops surrounding the missing point prevent a global potential."
    },
    {
      "heading": "Finding a potential",
      "text": "For an exact form in two variables, a potential can often be reconstructed by successive integration.\n\nConsider\n\n\\[\\omega=(2xy+3)\\,dx+(x^2+4y)\\,dy.\\]\n\nThe mixed-partial test gives\n\n\\[\\frac{\\partial}{\\partial y}(2xy+3)=2x = \\frac{\\partial}{\\partial x}(x^2+4y).\\]\n\nThe coefficients are defined on all of \\(\\mathbb{R}^2\\), so the form is exact. Integrating the first coefficient with respect to \\(x\\),\n\n\\[f(x,y)=x^2y+3x+g(y),\\]\n\nwhere the “constant of integration” may depend on \\(y\\). Differentiation with respect to \\(y\\) gives\n\n\\[\\frac{\\partial f}{\\partial y}=x^2+g'(y).\\]\n\nMatching the second coefficient requires \\(g'(y)=4y\\), hence\n\n\\[f(x,y)=x^2y+3x+2y^2+C.\\]\n\nOn a star-shaped domain, a potential can also be constructed by integrating along straight segments from a fixed point \\(p_0\\):\n\n\\[f(x)= \\int_0^1 \\sum_{i=1}^{n} a_i\\bigl(p_0+t(x-p_0)\\bigr) (x_i-p_{0,i})\\,dt.\\]\n\nFor a closed continuously differentiable one-form, differentiating this expression yields \\(df=\\omega\\)."
    },
    {
      "heading": "Exact differential equations and integrating factors",
      "text": "A first-order differential equation written as\n\n\\[M(x,y)\\,dx+N(x,y)\\,dy=0\\]\n\nis called exact when the form on the left is exact. If it equals \\(df\\), its solution curves lie in level sets\n\n\\[f(x,y)=C.\\]\n\nWhere \\(\\partial f/\\partial y\\neq0\\), such a level set can locally be expressed as a graph \\(y=y(x)\\), satisfying\n\n\\[\\frac{dy}{dx}=-\\frac{M}{N}.\\]\n\nCritical points of \\(f\\) require separate care because the level set may fail to be a regular curve there.\n\nA nonexact equation may become exact after multiplication by a nonvanishing integrating factor \\(\\mu(x,y)\\). The condition is\n\n\\[\\frac{\\partial(\\mu M)}{\\partial y} = \\frac{\\partial(\\mu N)}{\\partial x}.\\]\n\nIn useful special cases:\n\n- An integrating factor depending only on \\(x\\) can be found if\n\n\\[\\frac{M_y-N_x}{N}=g(x),\\]\n\ngiving\n\n\\[\\mu(x)=\\exp\\left(\\int g(x)\\,dx\\right).\\]\n- An integrating factor depending only on \\(y\\) can be found if\n\n\\[\\frac{N_x-M_y}{M}=h(y),\\]\n\ngiving\n\n\\[\\mu(y)=\\exp\\left(\\int h(y)\\,dy\\right).\\]\n\nThese formulas apply on regions where the denominators are nonzero. They enforce closedness; global exactness still depends on the domain."
    },
    {
      "heading": "Differential-geometric generalization",
      "text": "On a smooth manifold, an exact one-form is again a form \\(df\\) for a globally defined smooth function. At each point \\(p\\), \\(df_p\\) acts on the tangent space at \\(p\\). This formulation does not require a preferred coordinate system or a metric.\n\nThe terminology extends to higher-degree forms: a \\(k\\)-form \\(\\omega\\), with \\(k\\geq1\\), is exact if\n\n\\[\\omega=d\\eta\\]\n\nfor some \\((k-1)\\)-form \\(\\eta\\). Exact forms are closed because \\(d^2=0\\), but closed forms need not be exact.\n\nDe Rham cohomology measures this difference:\n\n\\[H_{\\mathrm{dR}}^k(M)= \\frac{\\{\\text{closed }k\\text{-forms on }M\\}} {\\{\\text{exact }k\\text{-forms on }M\\}}.\\]\n\nFor degree one, vanishing of this group means that every closed one-form has a global potential. The generalized Stokes' theorem relates these derivative conditions to integrals over boundaries."
    },
    {
      "heading": "Physical interpretation",
      "text": "In classical mechanics, a force with potential energy \\(V\\) satisfies\n\n\\[\\mathbf{F}=-\\nabla V.\\]\n\nIts work form is therefore exact:\n\n\\[\\mathbf{F}\\cdot d\\mathbf{r}=-dV.\\]\n\nWork between two points depends only on the endpoints. A local derivative test alone is insufficient to establish this globally if the accessible region has nontrivial topology.\n\nIn thermodynamics, a state function such as internal energy \\(U\\) has an exact differential \\(dU\\). Heat and work are process-dependent transfers and are conventionally written \\(\\delta Q\\) and \\(\\delta W\\), rather than as differentials of state functions. With work done by the system taken as positive, the first law is\n\n\\[dU=\\delta Q-\\delta W.\\]\n\nFor a reversible process,\n\n\\[dS=\\frac{\\delta Q_{\\mathrm{rev}}}{T},\\]\n\nwhere \\(S\\) is thermodynamic entropy and \\(T\\) is absolute temperature. Here \\(1/T\\) acts as an integrating factor for reversible heat, illustrating how a path-dependent transfer can be related to an exact state differential."
    }
  ],
  "links": [
    "differential-form",
    "open-set",
    "euclidean-space",
    "linear-functional",
    "antiderivative",
    "chain-rule",
    "fundamental-theorem-of-calculus",
    "line-integral",
    "vector-field",
    "gradient",
    "partial-derivative",
    "exterior-derivative",
    "poincare-lemma",
    "convex-set",
    "topology",
    "differential-equation",
    "manifold",
    "tangent-space",
    "de-rham-cohomology",
    "stokes-theorem",
    "classical-mechanics",
    "thermodynamics",
    "temperature"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}