0% found this document useful (0 votes)
60 views5 pages

Proof of L Hopitals Rule

The document provides a proof of L'Hopital's rule. It first states the rule and then provides Cauchy's mean value theorem, which is used in the proof. The proof considers two cases - when the limits are equal to zero and when they are infinite. It uses concepts like derivatives, limits, and the mean value theorem to show that the rule holds in both cases.

Uploaded by

Lohit Daksha
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
60 views5 pages

Proof of L Hopitals Rule

The document provides a proof of L'Hopital's rule. It first states the rule and then provides Cauchy's mean value theorem, which is used in the proof. The proof considers two cases - when the limits are equal to zero and when they are infinite. It uses concepts like derivatives, limits, and the mean value theorem to show that the rule holds in both cases.

Uploaded by

Lohit Daksha
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

PROOF OF L’HÔPITAL’S

RULE

n the text, we proved a special case of L’Hôpital’s Rule (Theorems 1 and 2 in LT Section
I 7.7 or ET Section 4.7). This supplement presents the complete proof.

THEOREM 1 Theorem L’hôpital’s Rule Assume that f (x) and g(x) are differen-
tiable on an open interval containing a and that

f (a) = g(a) = 0

Also assume that g  (x)  = 0 for x near but not equal to a. Then
f (x) f  (x)
lim = lim 
x→a g(x) x→a g (x)

provided that the limit on the right exists. The conclusion also holds if f (x) and g(x)
are differentiable for x near (but not equal to) a and
lim f (x) = ±∞ and lim g(x) = ±∞
x→a x→a

Furthermore, these limits may be replaced by one-sided limits.

The proof of L’Hôpital’s Rule makes use of the following generalization of the Mean
Value Theorem known as Cauchy’s Mean Value Theorem.

THEOREM 2 Cauchy’s Mean Value Theorem Assume that f (x) and g(x) are con-
tinuous on the closed interval [a, b] and differentiable on (a, b). Assume further that
g  (x)  = 0 on (a, b). Then there exists at least one value c in (a, b) such that
f  (c) f (b) − f (a)
=
g  (c) g(b) − g(a)

Proof First note that g(x) satisfies the hypotheses of the standard Mean Value Theorem
on [a, b]. Therefore, there exists r ∈ (a, b) such that
g(b) − g(a)
g  (r) =
b−a
By assumption, g  (r) is not equal to zero. It follows that that g(b) − g(a) is non-zero.
Now, just as in the proof of the standard Mean Value Theorem, we build a new function
h(x) to which Rolle’s Theorem applies. Set
f (b) − f (a)
h(x) = f (x) − g(x)
g(b) − g(a)
1
2 PROOF OF L’HÔPITAL’S RULE

We have not divided by zero since g(b) − g(a)  = 0. Furthermore, h(x) is continuous on
[a, b] and differentiable on (a, b) because this is true of both f (x) and g(x). Straightfor-
ward calculation yields
f (a)g(b) − f (b)g(a)
h(a) = h(b) =
g(b) − g(a)
Thus, the hypotheses of Rolle’s Theorem are satisfied and the conclusion holds, namely,
there exists c ∈ (a, b) such that h (c) = 0. In other words,
f (b) − f (a) 
h (c) = f  (c) − g (c) = 0
g(b) − g(a)
We may divide by g  (c) (which is nonzero by assumption) and rearrange to obtain the
desired equality:
f  (c) f (b) − f (a)

=
g (c) g(b) − g(a)

We shall carry out the proof of L’Hôpital’s Rule for right-hand limits. The proof
for left-hand limits is similar and the result for two-sided limits follows immediately by
combining the results for left and right-hand limits. We treat the two cases of L’Hôpital’s
Rule separately.

C A S E 1 f (a) = g(a) = 0.
Since g  (x) is non-zero near x = a, there is an interval (a, b) such that g  (x) is
positive or negative for x ∈ (a, b), and therefore, g(x) is either strictly increasing or
strictly decreasing on (a, b) (Figure 1). However, g(a) = 0, so g(x) itself is non-zero for
x ∈ (a, b). Thus for all x ∈ (a, b), the hypotheses of Cauchy’s Mean Value Theorem for
the interval [a, x] are satisfied and hence, there exists c ∈ (a, x) such that
a c x b f (x) − f (a) f  (c)
= 
FIGURE 1 Choose b so that g(x)  = 0 for g(x) − g(a) g (c)
x ∈ (a, b). Note that c → a+ as x → a+.
Since f (a) = g(a) = 0, this reduces to
f (x) f  (c)
= 
g(x) g (c)
As x tends to a from the right, the value c also tends to a from the right, and the desired
conclusion follows:
f (x) f  (c) f  (c)
lim = lim  = lim  =L
x→a+ g(x) x→a+ g (c) c→a+ g (c)

C A S E 2 Infinite Limits.
Now suppose that f (x) and g(x) are differentiable for x near (but not equal to) a and
satisfy
lim f (x) = ±∞ and lim g(x) = ±∞
x→a+ x→a+

We assume that
f  (x)
lim =L
x→a+ g  (x)
PROOF OF L’HÔPITAL’S RULE 3

By the formal definition of right-hand limits, for every  > 0, there exists δ > 0 such that
f  (x)
L− < <L+ for all a <x <a+δ 1
g  (x)

Set b = a + δ. Making δ smaller if necessary, we may assume that f (x) and g(x)
c
are differentiable on (a, b), continuous on [a, b] and g(x) = 0 for x ∈ (a, b). For every
a x b = a +δ
x ∈ (a, b), the hypotheses of Cauchy’s Mean Value Theorem are satisfied on (x, b), so
FIGURE 2 We choose b so that (1) holds there exists c in (x, b) such that (Figure 2)
for all x ∈ (a, b)
f (x) f (b)
f  (c) f (b) − f (x) g(x) − g(x)

= =
g (c) g(b) − g(x) 1− g(b)
g(x)

Multiplying by the denominator on the right, we rewrite this as follows


f  (c)  g(b)  f (x) f (b)
1 − = −
g  (c) g(x) g(x) g(x)

or

f  (c) f (x)  f (b) f  (c)g(b) 


= − − 
g  (c) g(x) g(x) g (c)g(x)
  
call this r(x)

Denote the indicated quantity on the right-hand by r(x). In other words,


f  (c) f (x)

= − r(x) 2
g (c) g(x)
Now apply (1) with x = c:
f  (c)
L− < <L+ 3
g  (c)
Together with Eq. (2), we obtain
f (x)
L −  + r(x) < < L +  + r(x) for all x ∈ (a, b) 4
g(x)
Keep in mind that b depends on the choice of . However, for fixed b, we claim that
r(x) tends to zero as x → a+. Indeed, the first term f (b)/g(x) tends to zero because
g(x) → ∞. Similarly, the second term f  (c)g(b)/g  (c)g(x) tends to zero because g(x) →
∞ and f  (c)/g  (c) remains bounded by (3). Having thus shown that r(x) tends to zero,
we may choose δ1 > 0 such that |r(x)| <  for all x ∈ (a, a + δ1 ). We may then apply
(4) to conclude that
f (x)
L − 2 < < L + 2 for all x ∈ (a, a + δ1 ) 5
g(x)
Since  is arbitrary, this suffices to prove that
f (x)
lim =L
x→a+ g(x)

as desired.
4 PROOF OF L’HÔPITAL’S RULE

C A S E 3 Limits as x → ∞.
We shall prove L’Hôpital’s Rule for limits as x → ∞. The case x → −∞ is similar.
Thus we assume there exists a number b such that g  (x)  = 0 for all x > b and that the
following limit exists:
f  (x)
L = lim
x→∞ g  (x)

We use the variable t = x −1 to convert limits as x → ∞ to one-sided limits as x → 0+.


Thus we have
f  (x) f  (t −1 )
lim = lim =L
t→∞ g  (x) t→0+ g  (t −1 )

By the Chain Rule,

[f (t −1 )] −t −2 f  (t −1 ) f  (t −1 )
= =
[g(t −1 )] −t −2 g  (t −1 ) g  (t −1 )
Therefore
[f (t −1 )] f  (t −1 )
lim = lim =L
t→0+ [g(t −1 )] t→0+ g  (t −1 )

Now we may apply L’Hôpital’s for one-sided limits as t → 0+ to reach the desired
conclusion:
f (x) f (t −1 ) [f (t −1 )]
lim = lim = lim =L
x→∞ g(x) t→0+ g(t −1 ) t→0+ [g(t −1 )]

Remark: In the statement and proof of L’Hôpital’s Rule, we have assumed that the limit
f  (x)
L = lim
x→a g  (x)
is a finite limit. This assumption is not necessary: L’Hôpital’s Rule remains valid if this
limit is infinite (i.e., L = ±∞). The above proofs may be easily modified to handle this
case. For example, suppose that L = ∞. In Case 1 (g(a) = f (a) = 0), we showed (in
the notation of the proof) that
f (x) f  (c)
= 
g(x) g (c)
Since c → a+ as x → a+, the ratio on the right tends to ∞ as x → a+ and the conclusion
follows. Similarly, the proof in Case 2 is valid, but the role of  > 0 is replaced by an
arbitrarily large number M > 0. Instead of (1), we have, for all M, the existence of δ > 0
such that
f  (x)
>M for all a < x < a + δ.
g  (x)
Instead of (4), we conclude that
f (x)
> M + r(x) for all x ∈ (a, b)
g(x)
PROOF OF L’HÔPITAL’S RULE 5

where b = a + δ. Since r(x) → 0 as x → a+, there exists δ1 > 0 such that |r(x)| < 12 M
for x ∈ (a, a + δ1 ) and therefore
f (x) 1
> M for all x ∈ (a, a + δ1 )
g(x) 2
f (x)
This proves that tends to ∞ as claimed.
g(x)

You might also like