LIMIT AND DERIVATIVE
MATHEMATICS EDUCATION DEPARTEMENT
FACULTY OF MATHEMATICS AND NATURAL SINCE
GANESHA UNIVERSITY OF EDUCATION
DEFINITION OF LIMIT
The concept of a "limit" is used to describe the value that a function or sequence "approaches" as the input or index approaches some value. The concept of limit allows one to, in a complete space, define a new point from a Cauchy sequence of previously defined points. Limits are essential to calculus (and mathematical analysis in general) and are used to define continuity, derivatives and integrals.
The concept of a limit of a sequence is further generalized to the concept of a limit of a topological net, and is closely related to limit and direct limit in category theory.
In formulas, limit is usually abbreviated as lim as in lim(an) = a or represented by the right arrow (→) as in an → a.
Whenever a point x is within δ units of c,f(x) is within ε units of L
For all x > S, f(x) is within ε of L
Suppose f(x) is a real-valued function and c is a real number. The expression
means that f(x) can be made to be as close to L as desired by making x sufficiently close to c. In that case, it can be stated that "the limit of fof x, as x approaches c, is L". Note that this statement can be true even if f(c) ≠ L. Indeed, the function f(x) need not even be defined at c.
For example, if
then f(1) is not defined, yet as x approaches 1, f(x) approaches 2:
⇒ undef ⇐
Thus, f(x) can be made arbitrarily close to the limit of 2 just by making x sufficiently close to 1.
Karl Weierstrass formalized the definition of the limit of a function into what became known as the (ε, δ)-definition of limit in the 19th century.
In addition to limits at finite values, functions can also have limits at infinity. For example, consider
§ f(100) = 1.9900
§ f(1000) = 1.9990
§ f(10000) = 1.9999
As x becomes extremely large, the value of f(x) approaches 2, and the value of f(x) can be made as close to 2 as one could wish just by pickingx sufficiently large. In this case, the limit of f(x) as x approaches infinity is 2. In mathematical notation,
Limit of a sequence
Consider the following sequence: 1.79, 1.799, 1.7999,... It can be observed that the numbers are "approaching" 1.8, the limit of the sequence.
Formally, suppose x1, x2, ... is a sequence of real numbers. It can be stated that the real number L is the limit of this sequence, namely:
For every real number ε > 0, there exists a natural number n0 such that for all n > n0, |xn − L| < ε.
Intuitively, this means that eventually all elements of the sequence get as close as needed to the limit, since the absolute value |xn − L| is the distance between xn and L. Not every sequence has a limit; if it does, it is called convergent, and if it does not, it is divergent. One can show that a convergent sequence has only one limit.
The limit of a sequence and the limit of a function are closely related. On one hand, the limit of a sequence is simply the limit at infinity of a function defined on natural numbers. On the other hand, a limit of a function f at x, if it exists, is the same as the limit of the sequence f(an) where an is any arbitrary sequence whose limit is x, and where an is never equal to x. Note that one such sequence would be x + 1/n.
Convergence and fixed point
A formal definition of convergence can be stated as follows. Suppose pn as n goes from 0 to is a sequence that converges to a fixed point p, with for all n. If positive constants λ and α exist with
then pn as n goes from 0 to converges to p of order α, with asymptotic error constant λ
Given a function f(x) = x with a fixed point p, there is a nice checklist for checking the convergence of p.
1) First check that p is indeed a fixed point:
f(p) = p
2) Check for linear convergence. Start by finding .
3) If it is found that there is something better than linear the expression should be checked for quadratic convergence. Start by finding If....
then there is quadratic convergence provided that is continuous
then there is something even better than quadratic convergence
does not exist
then there is convergence that is better than linear but still not quadratic
The derivative is a measure of how a function changes as its input changes. Loosely speaking, a derivative can be thought of as how much one quantity is changing in response to changes in some other quantity; for example, the derivative of the position of a moving object with respect to time is the object's instantaneous velocity (conversely, integrating a car's velocity over time yields the distance traveled).
The derivative of a function at a chosen input value describes the best linear approximation of the function near that input value. For a real-valued function of a single real variable, the derivative at a point equals the slope of the tangent line to the graph of the function at that point. In higher dimensions, the derivative of a function at a point is a linear transformation called the linearization. A closely related notion is the differential of a function.
The process of finding a derivative is called differentiation. The reverse process is called antidifferentiation. The fundamental theorem of calculus states that antidifferentiation is the same as integration. Differentiation and integration constitute the two fundamental operations in single-variable calculus.
Differentiation and the derivative
Differentiation is a method to compute the rate at which a dependent output y changes with respect to the change in the independent input x. This rate of change is called the derivative of y with respect to x. In more precise language, the dependence of y upon x means that y is a function of x. This functional relationship is often denoted y = ƒ(x), where ƒ denotes the function. If x and y are real numbers, and if the graph of y is plotted against x, the derivative measures the slope of this graph at each point.
The simplest case is when y is a linear function of x, meaning that the graph of y against x is a straight line. In this case, y = ƒ(x) = m x + b, for real numbers m and b, and the slope m is given by
where the symbol Δ (the uppercase form of the Greek letter Delta) is an abbreviation for "change in." This formula is true because
y + Δy = ƒ(x+ Δx) = m (x + Δx) + b = m x + b + m Δx = y + mΔx.
It follows that Δy = m Δx.
This gives an exact value for the slope of a straight line. If the function ƒ is not linear (i.e. its graph is not a straight line), however, then the change in y divided by the change in x varies: differentiation is a method to find an exact value for this rate of change at any given value of x.
Rate of change as a limiting value
Figure 1. The tangent line at (x, ƒ(x))
Figure 2. The secant to curve y= ƒ(x) determined by points (x, ƒ(x)) and (x+h, ƒ(x+h))
Figure 3. The tangent line as limit of secants
The idea, illustrated by Figures 1-3, is to compute the rate of change as the limiting value of the ratio of the differences Δy / Δx as Δx becomes infinitely small.
In Leibniz's notation, such an infinitesimal change in x is denoted by dx, and the derivative of y with respect to x is written
suggesting the ratio of two infinitesimal quantities. (The above expression is read as "the derivative of y with respect to x", "d y by d x", or "d y over d x". The oral form "d y d x" is often used conversationally, although it may lead to confusion.)
The most common approach to turn this intuitive idea into a precise definition uses limits, but there are other methods, such as non-standard analysis.
Definition via difference quotients
Let ƒ be a real valued function. In classical geometry, the tangent line to the graph of the function ƒ at a real number a was the unique line through the point (a, ƒ(a)) that did not meet the graph of ƒ transversally, meaning that the line did not pass straight through the graph. The derivative of y with respect to x at a is, geometrically, the slope of the tangent line to the graph of ƒ at a. The slope of the tangent line is very close to the slope of the line through (a, ƒ(a)) and a nearby point on the graph, for example (a + h, ƒ(a + h)). These lines are called secant lines. A value of h close to zero gives a good approximation to the slope of the tangent line, and smaller values (in absolute value) of h will, in general, give better approximations. The slope m of the secant line is the difference between the y values of these points divided by the difference between the x values, that is,
This expression is Newton's difference quotient. The derivative is the value of the difference quotient as the secant lines approach the tangent line. Formally, the derivative of the function ƒ at a is the limit
of the difference quotient as h approaches zero, if this limit exists. If the limit exists, then ƒ is differentiable at a. Here ƒ′ (a) is one of several common notations for the derivative (see below).
Equivalently, the derivative satisfies the property that
which has the intuitive interpretation (see Figure 1) that the tangent line to ƒ at a gives the best linear approximation
to ƒ near a (i.e., for small h). This interpretation is the easiest to generalize to other settings.
Substituting 0 for h in the difference quotient causes division by zero, so the slope of the tangent line cannot be found directly using this method. Instead, define Q(h) to be the difference quotient as a function of h:
Q(h) is the slope of the secant line between (a, ƒ(a)) and (a + h, ƒ(a + h)). If ƒ is a continuous function, meaning that its graph is an unbroken curve with no gaps, then Q is a continuous function away from the point h = 0. If the limit exists, meaning that there is a way of choosing a value for Q(0) that makes the graph of Q a continuous function, then the function ƒ is differentiable at the point a, and its derivative at a equals Q(0).
In practice, the existence of a continuous extension of the difference quotient Q(h) to h = 0 is shown by modifying the numerator to cancel h in the denominator. This process can be long and tedious for complicated functions, and many shortcuts are commonly used to simplify the process.
The squaring function ƒ(x) = x² is differentiable at x = 3, and its derivative there is 6. This result is established by calculating the limit as h approaches zero of the difference quotient of ƒ(3):
The last expression shows that the difference quotient equals 6 + h when h ≠ 0 and is undefined when h = 0, because of the definition of the difference quotient. However, the definition of the limit says the difference quotient does not need to be defined when h = 0. Hence the slope of the graph of the squaring function at the point (3, 9) is 6, and so its derivative at x = 3 is ƒ '(3) = 6.
More generally, a similar computation shows that the derivative of the squaring function at x = a is ƒ '(a) = 2a.
Continuity and differentiability
This function does not have a derivative at the marked point, as the function is not continuous there.
If y = ƒ(x) is differentiable at a, then ƒ must also be continuous at a. As an example, choose a point a and let ƒ be the step function that returns a value, say 1, for all x less than a, and returns a different value, say 10, for all x greater than or equal to a. ƒ cannot have a derivative at a. If h is negative, then a + h is on the low part of the step, so the secant line from a to a + h is very steep, and as h tends to zero the slope tends to infinity. If h is positive, then a + h is on the high part of the step, so the secant line from a to a + h has slope zero. Consequently the secant lines do not approach any single slope, so the limit of the difference quotient does not exist.
The absolute value function is continuous, but fails to be differentiable at x = 0 since the tangent slopes do not approach the same value from the left as they do from the right.
However, even if a function is continuous at a point, it may not be differentiable there. For example, the absolute value function y = |x| is continuous at x = 0, but it is not differentiable there. If h is positive, then the slope of the secant line from 0 to h is one, whereas if h is negative, then the slope of the secant line from 0 to h is negative one. This can be seen graphically as a "kink" or a "cusp" in the graph at x = 0. Even a function with a smooth graph is not differentiable at a point where its tangent is vertical: For instance, the function y = 3√x is not differentiable at x = 0.
In summary: for a function ƒ to have a derivative it is necessary for the function ƒ to be continuous, but continuity alone is not sufficient.
Most functions that occur in practice have derivatives at all points or at almost every point. Early in the history of calculus, many mathematicians assumed that a continuous function was differentiable at most points. Under mild conditions, for example if the function is a monotone function or a Lipschitz function, this is true. However, in 1872 Weierstrass found the first example of a function that is continuous everywhere but differentiable nowhere. This example is now known as the Weierstrass function. In 1931, Stefan Banach proved that the set of functions that have a derivative at some point is a meager set in the space of all continuous functions. Informally, this means that hardly any continuous functions have a derivative at even one point.
The derivative as a function
Let ƒ be a function that has a derivative at every point a in the domain of ƒ. Because every point a has a derivative, there is a function that sends the point a to the derivative of ƒ at a. This function is written f′(x) and is called the derivative function or the derivative of ƒ. The derivative of ƒ collects all the derivatives of ƒ at all the points in the domain of ƒ.
Sometimes ƒ has a derivative at most, but not all, points of its domain. The function whose value at a equals f′(a) whenever f′(a) is defined and elsewhere is undefined is also called the derivative of ƒ. It is still a function, but its domain is strictly smaller than the domain of ƒ.
Using this idea, differentiation becomes a function of functions: The derivative is an operator whose domain is the set of all functions that have derivatives at every point of their domain and whose range is a set of functions. If we denote this operator by D, then D(ƒ) is the function f′(x). Since D(ƒ) is a function, it can be evaluated at a point a. By the definition of the derivative function, D(ƒ)(a) = f′(a).
For comparison, consider the doubling function ƒ(x) =2x; ƒ is a real-valued function of a real number, meaning that it takes numbers as inputs and has numbers as outputs:
The operator D, however, is not defined on individual numbers. It is only defined on functions:
Because the output of D is a function, the output of D can be evaluated at a point. For instance, when D is applied to the squaring function,
D outputs the doubling function,
which we named ƒ(x). This output function can then be evaluated to get ƒ(1) = 2, ƒ(2) = 4, and so on.
Let ƒ be a differentiable function, and let f′(x) be its derivative. The derivative of f′(x) (if it has one) is written f′′(x) and is called the second derivative of ƒ. Similarly, the derivative of a second derivative, if it exists, is written f′′′(x) and is called the third derivative of ƒ. These repeated derivatives are called higher-order derivatives.
If x(t) represents the position of an object at time t, then the higher-order derivatives of x have physical interpretations. The second derivative of x is the derivative of x′(t), the velocity, and by definition this is the object's acceleration. The third derivative of x is defined to be the jerk, and the fourth derivative is defined to be the jounce.
A function ƒ need not have a derivative, for example, if it is not continuous. Similarly, even if ƒ does have a derivative, it may not have a second derivative. For example, let
Calculation shows that ƒ is a differentiable function whose derivative is
f′(x) is twice the absolute value function, and it does not have a derivative at zero. Similar examples show that a function can have k derivatives for any non-negative integer k but no (k + 1)-order derivative. A function that has k successive derivatives is called k times differentiable. If in addition the kth derivative is continuous, then the function is said to be of differentiability class Ck. (This is a stronger condition than having k derivatives. For an example, see differentiability class.) A function that has infinitely many derivatives is called infinitely differentiable or smooth.
On the real line, every polynomial function is infinitely differentiable. By standard differentiation rules, if a polynomial of degree n is differentiated n times, then it becomes a constant function. All of its subsequent derivatives are identically zero. In particular, they exist, so polynomials are smooth functions.
The derivatives of a function ƒ at a point x provide polynomial approximations to that function near x. For example, if ƒ is twice differentiable, then
in the sense that
If ƒ is infinitely differentiable, then this is the beginning of the Taylor series for ƒ.
A point where the second derivative of a function changes sign is called an inflection point. At an inflection point, the second derivative may be zero, as in the case of the inflection point x=0 of the function y=x3, or it may fail to exist, as in the case of the inflection point x=0 of the function y=x1/3. At an inflection point, a function switches from being a convex function to being a concave function or vice versa.
Notations for differentiation
The notation for derivatives introduced by Gottfried Leibniz is one of the earliest. It is still commonly used when the equation y = ƒ(x) is viewed as a functional relationship between dependent and independent variables. Then the first derivative is denoted by
and was once thought of as an infinitesimal quotient. Higher derivatives are expressed using the notation
for the nth derivative of y = ƒ(x) (with respect to x). These are abbreviations for multiple applications of the derivative operator. For example,
With Leibniz's notation, we can write the derivative of y at the point x = a in two different ways:
Leibniz's notation allows one to specify the variable for differentiation (in the denominator). This is especially relevant for partial differentiation. It also makes the chain rule easy to remember:
Sometimes referred to as prime notation, one of the most common modern notations for differentiation is due to Joseph-Louis Lagrange and uses the prime mark, so that the derivative of a function ƒ(x) is denoted ƒ′(x) or simply ƒ′. Similarly, the second and third derivatives are denoted
To denote the number of derivatives beyond this point, some authors use Roman numerals in superscript, whereas others place the number in parentheses:
The latter notation generalizes to yield the notation ƒ (n) for the nth derivative of ƒ — this notation is most useful when we wish to talk about the derivative as being a function itself, as in this case the Leibniz notation can become cumbersome.
Newton's notation for differentiation, also called the dot notation, places a dot over the function name to represent a time derivative. If y = ƒ(t), then
denote, respectively, the first and second derivatives of y with respect to t. This notation is used exclusively for time derivatives, meaning that the independent variable of the function represents time. It is very common in physics and in mathematical disciplines connected with physics such as differential equations. While the notation becomes unmanageable for high-order derivatives, in practice only very few derivatives are needed.
Euler's notation uses a differential operator D, which is applied to a function ƒ to give the first derivative Df. The second derivative is denoted D2ƒ, and the nth derivative is denoted Dnƒ.
If y = ƒ(x) is a dependent variable, then often the subscript x is attached to the D to clarify the independent variable x. Euler's notation is then written
although this subscript is often omitted when the variable x is understood, for instance when this is the only variable present in the expression.
Euler's notation is useful for stating and solving linear differential equations.
Computing the derivative
The derivative of a function can, in principle, be computed from the definition by considering the difference quotient, and computing its limit. In practice, once the derivatives of a few simple functions are known, the derivatives of other functions are more easily computed using rules for obtaining derivatives of more complicated functions from simpler ones.
Derivatives of elementary functions
Most derivative computations eventually require taking the derivative of some common functions. The following incomplete list gives some of the most frequently used functions of a single real variable and their derivatives.
- Derivatives of powers: if
where r is any real number, then
wherever this function is defined. For example, if f(x) = x1 / 4, then
and the derivative function is defined only for positive x, not for x = 0. When r = 0, this rule implies that f′(x) is zero for x ≠ 0, which is almost the constant rule (stated below).
- Exponential and logarithmic functions:
- Trigonometric functions:
- Inverse trigonometric functions:
Rules for finding the derivative
In many cases, complicated limit calculations by direct application of Newton's difference quotient can be avoided using differentiation rules. Some of the most basic rules are the following.
- Constant rule: if ƒ(x) is constant, then
- Sum rule:
for all functions ƒ and g and all real numbers a and b.
- Product rule:
for all functions ƒ and g.
- Quotient rule:
for all functions ƒ and g where g ≠ 0.
- Chain rule: If f(x) = h(g(x)), then
The derivative of
Here the second term was computed using the chain rule and third using the product rule. The known derivatives of the elementary functions x2, x4, sin(x), ln(x) and exp(x) = ex, as well as the constant 7, were also used.
Derivatives in higher dimensions
Derivatives of vector valued functions
A vector-valued function y(t) of a real variable sends real numbers to vectors in some vector space Rn. A vector-valued function can be split up into its coordinate functions y1(t), y2(t), …, yn(t), meaning that y(t) = (y1(t), ..., yn(t)). This includes, for example, parametric curves in R2 or R3. The coordinate functions are real valued functions, so the above definition of derivative applies to them. The derivative of y(t) is defined to be the vector, called the tangent vector, whose coordinates are the derivatives of the coordinate functions. That is,
if the limit exists. The subtraction in the numerator is subtraction of vectors, not scalars. If the derivative of y exists for every value of t, then y′ is another vector valued function.
If e1, …, en is the standard basis for Rn, then y(t) can also be written as y1(t)e1 + … + yn(t)en. If we assume that the derivative of a vector-valued function retains the linearity property, then the derivative of y(t) must be
because each of the basis vectors is a constant.
This generalization is useful, for example, if y(t) is the position vector of a particle at time t; then the derivative y′(t) is the velocity vector of the particle at time t.
Suppose that ƒ is a function that depends on more than one variable. For instance,
ƒ can be reinterpreted as a family of functions of one variable indexed by the other variables:
In other words, every value of x chooses a function, denoted fx, which is a function of one real number. That is,
Once a value of x is chosen, say a, then f(x,y) determines a function fa that sends y to a² + ay + y²:
In this expression, a is a constant, not a variable, so fa is a function of only one real variable. Consequently the definition of the derivative for a function of one variable applies:
The above procedure can be performed for any choice of a. Assembling the derivatives together into a function gives a function that describes the variation of ƒ in the y direction:
This is the partial derivative of ƒ with respect to y. Here ∂ is a rounded d called the partial derivative symbol. To distinguish it from the letter d, ∂ is sometimes pronounced "der", "del", or "partial" instead of "dee".
In general, the partial derivative of a function ƒ(x1, …, xn) in the direction xi at the point (a1 …, an) is defined to be:
In the above difference quotient, all the variables except xi are held fixed. That choice of fixed values determines a function of one variable
and, by definition,
In other words, the different choices of a index a family of one-variable functions just as in the example above. This expression also shows that the computation of partial derivatives reduces to the computation of one-variable derivatives.
An important example of a function of several variables is the case of a scalar-valued function ƒ(x1,...xn) on a domain in Euclidean space Rn (e.g., on R² or R³). In this case ƒ has a partial derivative ∂ƒ/∂xj with respect to each variable xj. At the point a, these partial derivatives define the vector
This vector is called the gradient of ƒ at a. If ƒ is differentiable at every point in some domain, then the gradient is a vector-valued function ∇ƒ that takes the point a to the vector ∇f(a). Consequently the gradient determines a vector field.
If ƒ is a real-valued function on Rn, then the partial derivatives of ƒ measure its variation in the direction of the coordinate axes. For example, if ƒ is a function of x and y, then its partial derivatives measure the variation in ƒ in the x direction and the y direction. They do not, however, directly measure the variation of ƒ in any other direction, such as along the diagonal line y = x. These are measured using directional derivatives. Choose a vector
The directional derivative of ƒ in the direction of v at the point x is the limit
In some cases it may be easier to compute or estimate the directional derivative after changing the length of the vector. Often this is done to turn the problem into the computation of a directional derivative in the direction of a unit vector. To see how this works, suppose that v = λu. Substitute h = k/λ into the difference quotient. The difference quotient becomes:
This is λ times the difference quotient for the directional derivative of f with respect to u. Furthermore, taking the limit as h tends to zero is the same as taking the limit as k tends to zero because h and k are multiples of each other. Therefore Dv(ƒ) = λDu(ƒ). Because of this rescaling property, directional derivatives are frequently considered only for unit vectors.
If all the partial derivatives of ƒ exist and are continuous at x, then they determine the directional derivative of ƒ in the direction v by the formula:
This is a consequence of the definition of the total derivative. It follows that the directional derivative is linear in v, meaning that Dv + w(ƒ) = Dv(ƒ) + Dw(ƒ).
The same definition also works when ƒ is a function with values in Rm. The above definition is applied to each component of the vectors. In this case, the directional derivative is a vector in Rm.
The total derivative, the total differential and the Jacobian
When ƒ is a function from an open subset of Rn to Rm, then the directional derivative of ƒ in a chosen direction is the best linear approximation to ƒ at that point and in that direction. But when n > 1, no single directional derivative can give a complete picture of the behavior of ƒ. The total derivative, also called the (total) differential, gives a complete picture by considering all directions at once. That is, for any vector v starting at a, the linear approximation formula holds:
Just like the single-variable derivative, ƒ ′(a) is chosen so that the error in this approximation is as small as possible.
If n and m are both one, then the derivative ƒ ′(a) is a number and the expression ƒ ′(a)v is the product of two numbers. But in higher dimensions, it is impossible for ƒ ′(a) to be a number. If it were a number, then ƒ ′(a)v would be a vector in Rn while the other terms would be vectors in Rm, and therefore the formula would not make sense. For the linear approximation formula to make sense, ƒ ′(a) must be a function that sends vectors in Rn to vectors in Rm, and ƒ ′(a)v must denote this function evaluated at v.
To determine what kind of function it is, notice that the linear approximation formula can be rewritten as
Notice that if we choose another vector w, then this approximate equation determines another approximate equation by substituting w for v. It determines a third approximate equation by substituting both w for v and a + v for a. By subtracting these two new equations, we get
If we assume that v is small and that the derivative varies continuously in a, then ƒ ′(a + v) is approximately equal to ƒ ′(a), and therefore the right-hand side is approximately zero. The left-hand side can be rewritten in a different way using the linear approximation formula with v + w substituted for v. The linear approximation formula implies:
This suggests that ƒ ′(a) is a linear transformation from the vector space Rn to the vector space Rm. In fact, it is possible to make this a precise derivation by measuring the error in the approximations. Assume that the error in these linear approximation formula is bounded by a constant times ||v||, where the constant is independent of v but depends continuously on a. Then, after adding an appropriate error term, all of the above approximate equalities can be rephrased as inequalities. In particular, ƒ ′(a) is a linear transformation up to a small error term. In the limit as v and w tend to zero, it must therefore be a linear transformation. Since we define the total derivative by taking a limit as v goes to zero, ƒ ′(a) must be a linear transformation.
In one variable, the fact that the derivative is the best linear approximation is expressed by the fact that it is the limit of difference quotients. However, the usual difference quotient does not make sense in higher dimensions because it is not usually possible to divide vectors. In particular, the numerator and denominator of the difference quotient are not even in the same vector space: The numerator lies in the codomain Rm while the denominator lies in the domain Rn. Furthermore, the derivative is a linear transformation, a different type of object from both the numerator and denominator. To make precise the idea that ƒ ′ (a) is the best linear approximation, it is necessary to adapt a different formula for the one-variable derivative in which these problems disappear. If ƒ : R → R, then the usual definition of the derivative may be manipulated to show that the derivative of ƒ at a is the unique number ƒ ′(a) such that
This is equivalent to
because the limit of a function tends to zero if and only if the limit of the absolute value of the function tends to zero. This last formula can be adapted to the many-variable situation by replacing the absolute values with norms.
The definition of the total derivative of ƒ at a, therefore, is that it is the unique linear transformation ƒ ′(a) : Rn → Rm such that
Here h is a vector in Rn, so the norm in the denominator is the standard length on Rn. However, ƒ′(a)h is a vector in Rm, and the norm in the numerator is the standard length on Rm. If v is a vector starting at a, then ƒ ′(a)v is called the pushforward of v by ƒ and is sometimes written ƒ*v.
If the total derivative exists at a, then all the partial derivatives and directional derivatives of ƒ exist at a, and for all v, ƒ ′(a)v is the directional derivative of ƒ in the direction v. If we write ƒ using coordinate functions, so that ƒ = (ƒ1, ƒ2, ..., ƒm), then the total derivative can be expressed using the partial derivatives as a matrix. This matrix is called the Jacobian matrix of ƒ at a:
The existence of the total derivative ƒ′(a) is strictly stronger than the existence of all the partial derivatives, but if the partial derivatives exist and are continuous, then the total derivative exists, is given by the Jacobian, and depends continuously on a.
The definition of the total derivative subsumes the definition of the derivative in one variable. That is, if ƒ is a real-valued function of a real variable, then the total derivative exists if and only if the usual derivative exists. The Jacobian matrix reduces to a 1×1 matrix whose only entry is the derivative ƒ′(x). This 1×1 matrix satisfies the property that ƒ(a + h) − ƒ(a) − ƒ ′(a)h is approximately zero, in other words that
Up to changing variables, this is the statement that the function is the best linear approximation to ƒ at a.
The total derivative of a function does not give another function in the same way as the one-variable case. This is because the total derivative of a multivariable function has to record much more information than the derivative of a single-variable function. Instead, the total derivative gives a function from the tangent bundle of the source to the tangent bundle of the target.
The natural analog of second, third, and higher-order total derivatives is not a linear transformation, is not a function on the tangent bundle, and is not built by repeatedly taking the total derivative. The analog of a higher-order derivative, called a jet, cannot be a linear transformation because higher-order derivatives reflect subtle geometric information, such as concavity, which cannot be described in terms of linear data such as vectors. It cannot be a function on the tangent bundle because the tangent bundle only has room for the base space and the directional derivatives. Because jets capture higher-order information, they take as arguments additional coordinates representing higher-order changes in direction. The space determined by these additional coordinates is called the jet bundle. The relation between the total derivative and the partial derivatives of a function is paralleled in the relation between the kth order jet of a function and its partial derivatives of order less than or equal to k.