34 DerivativeProofs
34 DerivativeProofs
(x1, y1)
f (x + h) − f (x)
) )
∆y = y1 − y2
( x ( t
f
f
=
=
y
y
∆x = x1 − x2 h
(x2, y2)
x, f (x)
∆y y2 − y 1 f (x1) − f (x2) f (x + h) − f (x)
= = =
∆x x2 − x 1 x1 − x 2 h
dy f (x2) − f (x1) f (x + h) − f (x)
= lim = lim = f 0(x)
dx x2→x1 x2 − x1 x+h→x h
d
y Leibniz Newton ẏ Lagrange y0
dx
Different Notations for Derivatives u, f (u)
(x1, y1)
) )
∆y = y1 − y2
t
f (u) − f (x)
( x (
f
f
=
=
y
y
∆x = x1 − x2 u−x
(x2, y2) x, f (x)
∆y y2 − y 1 f (x1) − f (x2) f (u) − f (x)
= = =
∆x x2 − x 1 x1 − x 2 u−x
dy f (x2) − f (x1) f (u) − f (x)
= lim = lim = f 0(x)
dx x2→x1 x2 − x 1 u→x u−x
d
y Leibniz Newton ẏ Lagrange y0
dx
f (x + h) − f (x)
A derivative’s definition, lim = f 0(x),
h→0 h
f (x + h) − f (x)
is equivalent to = f 0(x) + o(1),
h
Whenever we have
f (x + h) = a + bh + o(h)
then it must also be true that:
The value of f at x, f (x), equals a;
The derivative of f at x, f 0(x) equals b; and,
The graph of y = f (x + h) has the graph of
y = a + bh
as the line tangent to that curve at the point where h = 0.
The “Little Oh” Formula for a Derivative:
f (u) = f (x) + f 0(x)(u − x) + o(u − x)
Whenever we have
f (u) = a + b(u − x) + o(u − x)
then it must also be true that:
The value of f at x, f (x), equals a;
The derivative of f at x, f 0(x) equals b; and,
The graph of y = f (u) has the graph of
y = a + b(u − x)
as the line tangent to that curve at the point (x, a).
y
0
x + h, f (x) + f (x)h
x)
t −
) (
0( x o(h)
f
)+ 0
x + h, f (x) + f (x)h + o(h)
(x )
=
f f (t
y y=
f (x + h) = f (x) + f 0(x)h + o(h)
x, f (x)
x t
x+h
y
0
u, f (x) + f (x)(u − x)
x)
t −
) (
0 (x o(u − x)
f
) +
0
x u, f (x) + f (x)(u − x) + o(u − x)
f( f (t
)
=
y y=
x, f (x)
x u t
Compare with the “Little Oh” Formula for Continuity:
f (x + h) = f (x) + o(1)
y
( t)
f
= o(1)
y
y = f (x)
x t
x+h
Compare with the “Little Oh” Formula for Continuity:
f (u) = f (x) + o(1)
y
( t)
f
= o(1)
y
y = f (x)
x u t
Differentiation Examples in “Little-Oh” Notation
(x + h)2 = x2 + 2x · h + h2
= x2 + 2x · h + o(h)
(x + h)3 = x3 + 3x2 · h + 3x · h2 + h3
= x3 + 3x2 · h + 3x · o(h) + o(h)
= x3 + 3x2 · h + o(h)
Differentiation Examples in “Little-Oh” Notation
u2 = x2 + 2x · (u − x) + (u − x)2
= x2 + 2x · (u − x) + o(u − x)
!
1 1 1 1 1
h
− + − 2 h = − + 2
x+h x x x+h x x
x2 − (x + h)x + h(x + h)
=
(x + h)x2
x2 − x2 − hx + hx + h2
=
(x + h)x2
h2
= 2
= o(h)
(x + h)x
as h → 0.
d 1 1
Another way to show =− for x 6= 0
dx x x2
would be to simplify
!
1 1 1 1 1 u−x
− + − 2 (u − x) = − +
u x x u x x2
x2 − ux + u2 − ux
=
ux2
x2 − 2ux + u2
=
ux2
(u − x)2
= 2
= o(u − x)
ux
as u → x.
Theorem: Differentiability implies Continuity
1 1 1
= − (h) + o(h)
x+h x x2
1
= + o(1)
x
Theorem: Differentiability implies Continuity
u3 = x3 + 3x2(u − x) + o(u − x)
= x3 + o(1)
1 1 1
= − (u − x) + o(u − x)
u x x2
1
= + o(1)
x
Rules for Derivatives
d d d
(f (x) + g(x)) = f (x) + g(x)
dx dx dx
0
then f (x + h) · g(x + h) equals f (x) + f (x)h + o(h)
0
· g(x) + g (x)h + o(h) ,
d d d
(f (x) · g(x)) = f (x) · g(x) + f (x) · g(x)
dx dx dx
0
f (u) · g(u) = f (x) + f (x)(u − x) + o(u − x)
0
· g(x) + g (x)(u − x) + o(u − x) ,
which, when multiplied out, equals:
f (x)g(x) + f (x)g 0(x)h + f (x) · o(h)
+ f 0(x)g(x)h + f 0(x)g 0(x)h2
+ f 0(x)h · o(h)
+ g(x) · o(h)
+ g 0(x)h · o(h)
+ o(h) · o(h)
Proof:
1 1
If f (u) = for u 6= 0, then f (g(x)) =
u g(x)
1 1
f 0(u) = − for u 6= 0, and f 0(g(x)) = − .
u2 g 2(x)
d
The Chain Rule, f (g(x)) = f 0(g(x))g 0(x),
dx
d 1 1
then yields =− g 0(x).
dx g(x) g 2(x)
Quotient Rule:
d f (x) f 0(x)g(x) − f (x)g 0(x)
= 2
for g(x) 6= 0.
dx g(x) g (x)
d f (x) 1
d
= f (x) ·
dx g(x) dx g(x)
1 1
d d
= f (x) · + f (x) ·
dx g(x) dx g(x)
0
1 g (x)
0
= f (x) · + f (x) · − 2
g(x) g (x)
f 0(x) f (x)g 0(x)
= −
g(x) g 2(x)
f 0(x)g(x) f (x)g 0(x)
= 2
− 2
.
g (x) g (x)