In modern mathematics the differential of a function is the linear transformation associated
to each point in the domain of the function. This linear tranformation is given by the derivative .
For example if
f
:
R
→
R
{\displaystyle f:\mathbb{R} \to \mathbb{R}}
is given by
f
(
x
)
=
x
2
{\displaystyle f(x)=x^2}
the the derivative is
f
′
(
x
)
=
2
x
{\displaystyle f'(x)=2x}
. At
x
=
5
{\displaystyle x=5}
the function value is
f
(
5
)
=
25
{\displaystyle f(5)=25}
but
f
′
(
5
)
=
10
{\displaystyle f'(5)=10}
is the linear transformation
x
↦
10
x
{\displaystyle x\mapsto 10x}
.
Another if
F
(
x
,
y
)
=
x
2
+
3
y
{\displaystyle F(x,y)=x^2+3y}
at
x
=
a
,
y
=
b
{\displaystyle x=a,y=b}
it differential is the gradient
[
∂
F
(
a
,
b
)
∂
x
,
∂
F
(
a
,
b
)
∂
y
]
{\displaystyle \left[\frac{\part F(a,b)}{\part x},\frac{\part F (a,b)}{\part y}\right]}
and determines the linear tranformation
[
∂
F
(
a
,
b
)
∂
x
,
∂
F
(
a
,
b
)
∂
y
]
:
R
2
→
R
{\displaystyle \left[\frac{\part F(a,b)}{\part x},\frac{\part F(a,b)}{\part y}\right]:\R^2\to\R}
given by
(
x
,
y
)
↦
∂
F
(
a
,
b
)
∂
x
x
+
∂
F
(
a
,
b
)
∂
y
y
{\displaystyle (x,y)\mapsto\frac{\part F(a,b)}{\part x}x+\frac{\part F(a,b)}{\part y}y}
For a vector function
F
:
R
n
→
R
m
{\displaystyle F:\R^n\to\R^m}
let us ilustrate with another beispiel:
Suppose that
[
v
w
]
↦
[
5
v
+
w
v
2
−
v
+
8
w
]
{\displaystyle
\begin{bmatrix}v\\w\end{bmatrix}\mapsto\begin{bmatrix}5v+w\\v^2\\-v+8w\end{bmatrix}
}
then
[
5
1
2
v
0
−
1
8
]
{\displaystyle \begin{bmatrix}5&1\\2v&0\\-1&8\end{bmatrix}}
is the Jacobian . So at
v
=
2
,
w
=
3
{\displaystyle v=2,w=3}
the differential is the map
[
v
w
]
↦
[
5
1
4
0
−
1
8
]
[
v
w
]
=
[
5
v
+
w
4
v
−
v
+
8
w
]
{\displaystyle
\begin{bmatrix}v\\w\end{bmatrix}\mapsto
\begin{bmatrix}5&1\\4&0\\-1&8\end{bmatrix}
\begin{bmatrix}v\\w\end{bmatrix}
=\begin{bmatrix}5v+w\\4v\\-v+8w\end{bmatrix}
}