In

^{T} denotes the transpose Jacobian matrix.
For the second form of the chain rule, suppose that is a real valued function on a subset of , and that is differentiable at the point . Then
:$\backslash nabla\; (h\backslash circ\; f)(a)\; =\; h\text{'}\backslash big(f(a)\backslash big)\backslash nabla\; f(a).$

vector calculus
Vector calculus, or vector analysis, is concerned with derivative, differentiation and integral, integration of vector fields, primarily in 3-dimensional Euclidean space \mathbb^3. The term "vector calculus" is sometimes used as a synonym for ...

, the gradient of a scalar-valued differentiable function
In mathematics
Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus, change (mathematical analysis, analysis). It ...

of several variables is the vector field
In vector calculus and physics, a vector field is an assignment of a vector to each point in a subset of space. For instance, a vector field in the plane can be visualised as a collection of arrows with a given magnitude and direction, each att ...

(or vector-valued function
A vector-valued function, also referred to as a vector function, is a function (mathematics), mathematical function of one or more variables whose range of a function, range is a set of multidimensional Euclidean vector, vectors or infinite-dimensi ...

) $\backslash nabla\; f$ whose value at a point $p$ is the vector
Vector may refer to:
Biology
*Vector (epidemiology)
In epidemiology
Epidemiology is the study and analysis of the distribution (who, when, and where), patterns and risk factor, determinants of health and disease conditions in defined pop ...

whose components are the partial derivative
In mathematics
Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus, change (mathematical analysis, analysis). It ...

s of $f$ at $p$. That is, for $f\; \backslash colon\; \backslash R^n\; \backslash to\; \backslash R$, its gradient $\backslash nabla\; f\; \backslash colon\; \backslash R^n\; \backslash to\; \backslash R^n$ is defined at the point $p\; =\; (x\_1,\backslash ldots,x\_n)$ in ''n-''dimensional space as the vector:
:$\backslash nabla\; f(p)\; =\; \backslash begin\; \backslash frac(p)\; \backslash \backslash \; \backslash vdots\; \backslash \backslash \; \backslash frac(p)\; \backslash end.$
The nabla symbol
∇
The nabla symbol
The nabla is a triangular symbol resembling an inverted Greek delta:Indeed, it is called ''anadelta'' ( ανάδελτα) in Modern Greek
Modern Greek (, , or , ''Kiní Neoellinikí Glóssa''), generally referred to by ...

$\backslash nabla$, written as an upside-down triangle and pronounced "del", denotes the .
The gradient vector can be interpreted as the "direction and rate of fastest increase". If the gradient of a function is non-zero at a point , the direction of the gradient is the direction in which the function increases most quickly from , and the magnitude
Magnitude may refer to:
Mathematics
*Euclidean vector, a quantity defined by both its magnitude and its direction
*Magnitude (mathematics), the relative size of an object
*Norm (mathematics), a term for the size or length of a vector
*Order of ...

of the gradient is the rate of increase in that direction, the greatest absoluteAbsolute may refer to:
Companies
* Absolute Entertainment, a video game publisher
* Absolute Radio, (formerly Virgin Radio), independent national radio station in the UK
* Absolute Software Corporation, specializes in security and data risk managem ...

directional derivative. Further, the gradient is the zero vector at a point if and only if it is a stationary point
In mathematics
Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus, change (mathematical analysis, analysis). I ...

(where the derivative vanishes). The gradient thus plays a fundamental role in optimization theory
Nelder-Mead minimum search of Simionescu's function. Simplex vertices are ordered by their values, with 1 having the lowest ( best) value., alt=
Mathematical optimization (alternatively spelled ''optimisation'') or mathematical programming i ...

, where it is used to maximize a function by gradient ascent
Gradient descent (also often called steepest descent) is a :First order methods, first-order Iterative algorithm, iterative Mathematical optimization, optimization algorithm for finding a local minimum of a differentiable function. The idea is to ...

.
The gradient is dual to the total derivative
In mathematics
Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus, change (mathematical analysis, analysis). I ...

$df$: the value of the gradient at a point is a tangent vector
:''For a more general — but much more technical — treatment of tangent vectors, see tangent space.''
In mathematics, a tangent vector is a Vector (geometry), vector that is tangent to a curve or Surface (mathematics), surface at a given point. T ...

– a vector at each point; while the value of the derivative at a point is a ''co''tangent vector – a linear function on vectors. They are related in that the dot product of the gradient of at a point with another tangent vector equals the directional derivative of at of the function along ; that is, $\backslash nabla\; f(p)\; \backslash cdot\; \backslash mathbf\; v\; =\; \backslash frac(p)\; =\; df\_(\backslash mathbf)$.
The gradient admits multiple generalizations to more general functions on manifolds; see .
Motivation

Consider a room where the temperature is given by a scalar field, , so at each point the temperature is , independent of time. At each point in the room, the gradient of at that point will show the direction in which the temperature rises most quickly, moving away from . The magnitude of the gradient will determine how fast the temperature rises in that direction. Consider a surface whose height above sea level at point is . The gradient of at a point is a plane vector pointing in the direction of the steepest slope or Grade (slope), grade at that point. The steepness of the slope at that point is given by the magnitude of the gradient vector. The gradient can also be used to measure how a scalar field changes in other directions, rather than just the direction of greatest change, by taking a dot product. Suppose that the steepest slope on a hill is 40%. A road going directly uphill has slope 40%, but a road going around the hill at an angle will have a shallower slope. For example, if the road is at a 60° angle from the uphill direction (when both directions are projected onto the horizontal plane), then the slope along the road will be the dot product between the gradient vector and a unit vector along the road, namely 40% times the cosine of 60°, or 20%. More generally, if the hill height function is differentiable function, differentiable, then the gradient of dot product, dotted with a unit vector gives the slope of the hill in the direction of the vector, the directional derivative of along the unit vector.Notation

The gradient of a function $f$ at point $a$ is usually written as $\backslash nabla\; f\; (a)$. It may also be denoted by any of the following: * $\backslash vec\; f\; (a)$ : to emphasize the vector nature of the result. * * $\backslash left.\; \backslash frac\backslash \_$ * $\backslash partial\_i\; f$ and $f\_$ : Einstein notation.Definition

The gradient (or gradient vector field) of a scalar function is denoted or where (nabla symbol, nabla) denotes the vector differential operator, del. The notation is also commonly used to represent the gradient. The gradient of is defined as the unique vector field whose dot product with any Euclidean vector, vector at each point is the directional derivative of along . That is, :$\backslash big(\backslash nabla\; f(x)\backslash big)\backslash cdot\; \backslash mathbf\; =\; D\_f(x).$ Formally, the gradient is ''dual'' to the derivative; see #Derivative, relationship with derivative. When a function also depends on a parameter such as time, the gradient often refers simply to the vector of its spatial derivatives only (see Spatial gradient). The magnitude and direction of the gradient vector are Invariant (mathematics), independent of the particular Coordinate system, coordinate representation.Cartesian coordinates

In the three-dimensional Cartesian coordinate system with a Euclidean metric, the gradient, if it exists, is given by: :$\backslash nabla\; f\; =\; \backslash frac\; \backslash mathbf\; +\; \backslash frac\; \backslash mathbf\; +\; \backslash frac\; \backslash mathbf,$ where , , are the standard basis, standard unit vectors in the directions of the , and coordinates, respectively. For example, the gradient of the function :$f(x,y,z)=\; 2x+3y^2-\backslash sin(z)$ is :$\backslash nabla\; f\; =\; 2\backslash mathbf+\; 6y\backslash mathbf\; -\backslash cos(z)\backslash mathbf.$ In some applications it is customary to represent the gradient as a row vector or column vector of its components in a rectangular coordinate system; this article follows the convention of the gradient being a column vector, while the derivative is a row vector.Cylindrical and spherical coordinates

In cylindrical coordinate system#Definition, cylindrical coordinates with a Euclidean metric, the gradient is given by:. :$\backslash nabla\; f(\backslash rho,\; \backslash varphi,\; z)\; =\; \backslash frac\backslash mathbf\_\backslash rho\; +\; \backslash frac\backslash frac\backslash mathbf\_\backslash varphi\; +\; \backslash frac\backslash mathbf\_z,$ where is the axial distance, is the azimuthal or azimuth angle, is the axial coordinate, and , and are unit vectors pointing along the coordinate directions. In spherical coordinate system#Definition, spherical coordinates, the gradient is given by: :$\backslash nabla\; f(r,\; \backslash theta,\; \backslash varphi)\; =\; \backslash frac\backslash mathbf\_r\; +\; \backslash frac\backslash frac\backslash mathbf\_\backslash theta\; +\; \backslash frac\backslash frac\backslash mathbf\_\backslash varphi,$ where is the radial distance, is the azimuthal angle and is the polar angle, and , and are again local unit vectors pointing in the coordinate directions (that is, the normalized Curvilinear coordinates#Covariant and contravariant bases, covariant basis). For the gradient in other orthogonal coordinate systems, see Orthogonal coordinates#Differential operators in three dimensions, Orthogonal coordinates (Differential operators in three dimensions).General coordinates

We consider Curvilinear coordinates, general coordinates, which we write as , where is the number of dimensions of the domain. Here, the upper index refers to the position in the list of the coordinate or component, so refers to the second component—not the quantity squared. The index variable refers to an arbitrary element . Using Einstein notation, the gradient can then be written as: :$\backslash nabla\; f\; =\; \backslash fracg^\; \backslash mathbf\_j$ ( Note that its Dual space, dual is $\backslash mathrmf=\; \backslash frac\backslash mathbf^i$ ), where $\backslash mathbf\_i\; =\; \backslash partial\; \backslash mathbf/\backslash partial\; x^i$ and $\backslash mathbf^i\; =\; \backslash mathrmx^i$ refer to the unnormalized local Curvilinear coordinates#Covariant and contravariant bases, covariant and contravariant bases respectively, $g^$ is the Metric tensor#Inverse metric, inverse metric tensor, and the Einstein summation convention implies summation over ''i'' and ''j''. If the coordinates are orthogonal we can easily express the gradient (and the Differential form, differential) in terms of the normalized bases, which we refer to as $\backslash hat\_i$ and $\backslash hat^i$, using the scale factors (also known as Lamé coefficients) $h\_i=\; \backslash lVert\; \backslash mathbf\_i\; \backslash rVert\; =\; 1\backslash ,\; /\; \backslash lVert\; \backslash mathbf^i\; \backslash rVert$ : :$\backslash nabla\; f\; =\; \backslash sum\_^n\; \backslash ,\; \backslash frac\backslash frac\backslash mathbf\_i$ ( and $\backslash mathrmf\; =\; \backslash sum\_^n\; \backslash ,\; \backslash frac\backslash frac\backslash mathbf^i$ ), where we cannot use Einstein notation, since it is impossible to avoid the repetition of more than two indices. Despite the use of upper and lower indices, $\backslash mathbf\_i$, $\backslash mathbf^i$, and $h\_i$ are neither contravariant nor covariant. The latter expression evaluates to the expressions given above for cylindrical and spherical coordinates.Relationship with derivative

Relationship with total derivative

The gradient is closely related to thetotal derivative
In mathematics
Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus, change (mathematical analysis, analysis). I ...

(total differential) $df$: they are transpose (Transpose of a linear map, dual) to each other. Using the convention that vectors in $\backslash R^n$ are represented by column vectors, and that covectors (linear maps $\backslash R^n\; \backslash to\; \backslash R$) are represented by row vectors, the gradient $\backslash nabla\; f$ and the derivative $df$ are expressed as a column and row vector, respectively, with the same components, but transpose of each other:
:$\backslash nabla\; f(p)\; =\; \backslash begin\backslash frac(p)\; \backslash \backslash \; \backslash vdots\; \backslash \backslash \; \backslash frac(p)\; \backslash end\; ;$
:$df\_p\; =\; \backslash begin\backslash frac(p)\; \&\; \backslash cdots\; \&\; \backslash frac(p)\; \backslash end\; .$
While these both have the same components, they differ in what kind of mathematical object they represent: at each point, the derivative is a cotangent vector, a linear form (covector) which expresses how much the (scalar) output changes for a given infinitesimal change in (vector) input, while at each point, the gradient is a tangent vector
:''For a more general — but much more technical — treatment of tangent vectors, see tangent space.''
In mathematics, a tangent vector is a Vector (geometry), vector that is tangent to a curve or Surface (mathematics), surface at a given point. T ...

, which represents an infinitesimal change in (vector) input. In symbols, the gradient is an element of the tangent space at a point, $\backslash nabla\; f(p)\; \backslash in\; T\_p\; \backslash R^n$, while the derivative is a map from the tangent space to the real numbers, $df\_p\; \backslash colon\; T\_p\; \backslash R^n\; \backslash to\; \backslash R$. The tangent spaces at each point of $\backslash R^n$ can be "naturally" identified with the vector space $\backslash R^n$ itself, and similarly the cotangent space at each point can be naturally identified with the dual vector space $(\backslash R^n)^*$ of covectors; thus the value of the gradient at a point can be thought of a vector in the original $\backslash R^n$, not just as a tangent vector.
Computationally, given a tangent vector, the vector can be ''multiplied'' by the derivative (as matrices), which is equal to taking the dot product with the gradient:
:$(df\_p)(v)\; =\; \backslash begin\backslash frac(p)\; \&\; \backslash cdots\; \&\; \backslash frac(p)\; \backslash end\; \backslash beginv\_1\; \backslash \backslash \; \backslash vdots\; \backslash \backslash \; v\_n\backslash end\; =\; \backslash sum\_^n\; \backslash frac(p)\; v\_i\; =\; \backslash begin\backslash frac(p)\; \backslash \backslash \; \backslash vdots\; \backslash \backslash \; \backslash frac(p)\; \backslash end\; \backslash cdot\; \backslash beginv\_1\; \backslash \backslash \; \backslash vdots\; \backslash \backslash \; v\_n\backslash end\; =\; \backslash nabla\; f(p)\; \backslash cdot\; v$
Differential or (exterior) derivative

The best linear approximation to a differentiable function :$f\; \backslash colon\; \backslash R^n\; \backslash to\; \backslash R$ at a point in is a linear map from to which is often denoted by or and called the differential (calculus), differential ortotal derivative
In mathematics
Mathematics (from Ancient Greek, Greek: ) includes the study of such topics as quantity (number theory), mathematical structure, structure (algebra), space (geometry), and calculus, change (mathematical analysis, analysis). I ...

of at . The function , which maps to , is called the total differential or exterior derivative of and is an example of a differential 1-form.
Much as the derivative of a function of a single variable represents the slope of the tangent to the graph of a function, graph of the function, the directional derivative of a function in several variables represents the slope of the tangent hyperplane in the direction of the vector.
The gradient is related to the differential by the formula
:$(\backslash nabla\; f)\_x\backslash cdot\; v\; =\; df\_x(v)$
for any , where $\backslash cdot$ is the dot product: taking the dot product of a vector with the gradient is the same as taking the directional derivative along the vector.
If is viewed as the space of (dimension ) column vectors (of real numbers), then one can regard as the row vector with components
:$\backslash left(\; \backslash frac,\; \backslash dots,\; \backslash frac\backslash right),$
so that is given by matrix multiplication. Assuming the standard Euclidean metric on , the gradient is then the corresponding column vector, that is,
:$(\backslash nabla\; f)\_i\; =\; df^\backslash mathsf\_i.$
Linear approximation to a function

The best linear approximation to a function can be expressed in terms of the gradient, rather than the derivative. The gradient of a function (mathematics), function from the Euclidean space to at any particular point in characterizes the best linear approximation to at . The approximation is as follows: :$f(x)\; \backslash approx\; f(x\_0)\; +\; (\backslash nabla\; f)\_\backslash cdot(x-x\_0)$ for close to , where is the gradient of computed at , and the dot denotes the dot product on . This equation is equivalent to the first two terms in the Taylor series#Taylor series in several variables, multivariable Taylor series expansion of at .Relationship with Fréchet derivative

Let be an open set in . If the function is differentiable, then the differential of is the Fréchet derivative of . Thus is a function from to the space such that :$\backslash lim\_\; \backslash frac\; =\; 0,$ where · is the dot product. As a consequence, the usual properties of the derivative hold for the gradient, though the gradient is not a derivative itself, but rather dual to the derivative: ;Linearity The gradient is linear in the sense that if and are two real-valued functions differentiable at the point , and and are two constants, then is differentiable at , and moreover :$\backslash nabla\backslash left(\backslash alpha\; f+\backslash beta\; g\backslash right)(a)\; =\; \backslash alpha\; \backslash nabla\; f(a)\; +\; \backslash beta\backslash nabla\; g\; (a).$ ;Product rule If and are real-valued functions differentiable at a point , then the product rule asserts that the product is differentiable at , and :$\backslash nabla\; (fg)(a)\; =\; f(a)\backslash nabla\; g(a)\; +\; g(a)\backslash nabla\; f(a).$ ;Chain rule Suppose that is a real-valued function defined on a subset of , and that is differentiable at a point . There are two forms of the chain rule applying to the gradient. First, suppose that the function is a parametric curve; that is, a function maps a subset into . If is differentiable at a point such that , then :$(f\backslash circ\; g)\text{'}(c)\; =\; \backslash nabla\; f(a)\backslash cdot\; g\text{'}(c),$ where ∘ is the composition operator: . More generally, if instead , then the following holds: :$\backslash nabla\; (f\backslash circ\; g)(c)\; =\; \backslash big(Dg(c)\backslash big)^\backslash mathsf\; \backslash big(\backslash nabla\; f(a)\backslash big),$ whereFurther properties and applications

Level sets

A level surface, or isosurface, is the set of all points where some function has a given value. If is differentiable, then the dot product of the gradient at a point with a vector gives the directional derivative of at in the direction . It follows that in this case the gradient of is orthogonal to the level sets of . For example, a level surface in three-dimensional space is defined by an equation of the form . The gradient of is then normal to the surface. More generally, any embedded submanifold, embedded hypersurface in a Riemannian manifold can be cut out by an equation of the form such that is nowhere zero. The gradient of is then normal to the hypersurface. Similarly, an affine algebraic variety, affine algebraic hypersurface may be defined by an equation , where is a polynomial. The gradient of is zero at a singular point of the hypersurface (this is the definition of a singular point). At a non-singular point, it is a nonzero normal vector.Conservative vector fields and the gradient theorem

The gradient of a function is called a gradient field. A (continuous) gradient field is always a conservative vector field: its line integral along any path depends only on the endpoints of the path, and can be evaluated by the gradient theorem (the fundamental theorem of calculus for line integrals). Conversely, a (continuous) conservative vector field is always the gradient of a function.Generalizations

Jacobian

The Jacobian matrix is the generalization of the gradient for vector-valued functions of several variables and differentiable maps between Euclidean spaces or, more generally, manifolds. A further generalization for a function between Banach spaces is the Fréchet derivative. Suppose is a function such that each of its first-order partial derivatives exist on . Then the Jacobian matrix of is defined to be an matrix, denoted by $\backslash mathbf\_\backslash mathbb(\backslash mathbb)$ or simply $\backslash mathbf$. The th entry is $\backslash mathbf\; J\_\; =\; \backslash frac$. Explicitly : $\backslash mathbf\; J\; =\; \backslash begin\; \backslash dfrac\; \&\; \backslash cdots\; \&\; \backslash dfrac\; \backslash end\; =\; \backslash begin\; \backslash nabla^\backslash mathsf\; f\_1\; \backslash \backslash \; \backslash vdots\; \backslash \backslash \; \backslash nabla^\backslash mathsf\; f\_m\; \backslash end\; =\; \backslash begin\; \backslash dfrac\; \&\; \backslash cdots\; \&\; \backslash dfrac\backslash \backslash \; \backslash vdots\; \&\; \backslash ddots\; \&\; \backslash vdots\backslash \backslash \; \backslash dfrac\; \&\; \backslash cdots\; \&\; \backslash dfrac\; \backslash end.$Gradient of a vector field

Since the total derivative of a vector field is a linear mapping from vectors to vectors, it is a tensor quantity. In rectangular coordinates, the gradient of a vector field is defined by: :$\backslash nabla\; \backslash mathbf=g^\backslash frac\; \backslash mathbf\_i\; \backslash otimes\; \backslash mathbf\_k,$ (where the Einstein summation notation is used and the tensor product of the vectors and is a dyadic tensor of type (2,0)). Overall, this expression equals the transpose of the Jacobian matrix: :$\backslash frac\; =\; \backslash frac.$ In curvilinear coordinates, or more generally on a curved Riemannian manifold, manifold, the gradient involves Christoffel symbols: :$\backslash nabla\; \backslash mathbf=g^\backslash left(\backslash frac+\_f^l\backslash right)\; \backslash mathbf\_i\; \backslash otimes\; \backslash mathbf\_k,$ where are the components of the inverse metric tensor and the are the coordinate basis vectors. Expressed more invariantly, the gradient of a vector field can be defined by the Levi-Civita connection and metric tensor:. :$\backslash nabla^a\; f^b\; =\; g^\; \backslash nabla\_c\; f^b\; ,$ where is the connection.Riemannian manifolds

For any smooth function on a Riemannian manifold , the gradient of is the vector field such that for any vector field , :$g(\backslash nabla\; f,\; X)\; =\; \backslash partial\_X\; f,$ that is, :$g\_x\backslash big((\backslash nabla\; f)\_x,\; X\_x\; \backslash big)\; =\; (\backslash partial\_X\; f)\; (x),$ where denotes the inner product of tangent vectors at defined by the metric and is the function that takes any point to the directional derivative of in the direction , evaluated at . In other words, in a coordinate chart from an open subset of to an open subset of , is given by: :$\backslash sum\_^n\; X^\; \backslash big(\backslash varphi(x)\backslash big)\; \backslash frac(f\; \backslash circ\; \backslash varphi^)\; \backslash Bigg,\; \_,$ where denotes the th component of in this coordinate chart. So, the local form of the gradient takes the form: :$\backslash nabla\; f\; =\; g^\; \backslash frac\; \_i\; .$ Generalizing the case , the gradient of a function is related to its exterior derivative, since :$(\backslash partial\_X\; f)\; (x)\; =\; (df)\_x(X\_x)\; .$ More precisely, the gradient is the vector field associated to the differential 1-form using the musical isomorphism :$\backslash sharp=\backslash sharp^g\backslash colon\; T^*M\backslash to\; TM$ (called "sharp") defined by the metric . The relation between the exterior derivative and the gradient of a function on is a special case of this in which the metric is the flat metric given by the dot product.See also

* Curl (mathematics), Curl * Divergence * Four-gradient * Hessian matrix * Skew gradientNotes

References

* * * * * * * * * * * *Further reading

*External links

* * . * {{Calculus topics Differential operators Differential calculus Generalizations of the derivative Linear operators in calculus Vector calculus Rates