Proof of L'Hôpital's rule

-

To prove L'Hôpital's rule, we first prove Cauchy's mean value theorem that generalizes the mean value theorem provided earlier.



See also: Mean Value Theorem

Theorem (Cauchy's Mean Value Theorem)

Let \(f(x)\) and \(g(x)\) be functions that are continuous on \([a,b]\) and differentiable on \((a,b)\). Suppose that \(g'(x) \neq 0\) for all \(x \in (a,b)\).  Then, there exists a \(c\in (a,b)\) such that

\[\frac{f'(c)}{g'(c)} = \frac{f(b) - f(a)}{g(b) - g(a)}.\]

Proof. Note that the function \(g(x)\) satisfies the conditions of the mean value theorem. Thus, there exists a \(d\in(a,b)\) such that

\[g'(d) = \frac{g(b) - g(a)}{b - a}.\]

Since \(g'(d) \neq 0\) by assumption, it follows that \(g(b) - g(a) \neq 0\). Now let us define

\[h(x) = f(x) - \frac{f(b) - f(a)}{g(b) - g(a)}g(x).\]

Then \(h(x)\) is continuous on \([a,b]\) and differentiable on \((a,b)\). Moreover,

\[h(a) = \frac{f(a)g(b) - f(b)g(a)}{g(b) - g(a)} = h(b).\]

Therefore, by Rolle's theorem (Theorem 10.6), there exists a \(c\in (a,b)\) such that \(h'(c) = 0\). Then

\[h'(c) = f'(c) - \frac{f(b) - f(a)}{g(b) - g(a)}g'(c) = 0\]

so that, since \(g'(c) \neq 0\),

\[\frac{f'(c)}{g'(c)} = \frac{f(b) - f(a)}{g(b) - g(a)}.\]

■

Theorem L'Hôpital's rule (1)

Let \(f(x)\) and \(g(x)\) be differentiable functions on \((a,b)\). Assume that the following conditions are satisfied.

  1. \[\lim_{x\to a+0}f(x) = \lim_{x\to a+0}g(x) = 0.\]
  2. For all \(x \in (a,b)\), \(g'(x) \neq 0\).
  3. The right limit \[\lim_{x\to a+0}\frac{f'(x)}{g'(x)}\] exists.
Then, the right limit \(\lim_{x\to a+0}\frac{f(x)}{g(x)}\) exists and
\[\lim_{x\to a+0}\frac{f(x)}{g(x)} = \lim_{x\to a+0}\frac{f'(x)}{g'(x)}.\]
Furthermore, Condition 1 may be replaced with
\[ \lim_{x\to a+0}f(x) = \lim_{x\to a+0}g(x) = \pm \infty. \]
Proof. For convenience we set \(f(a) = g(a) = 0\) so that \(f(x)\) and \(g(x)\) are defined on \([a, b)\). Assume conditions 1, 2, and 3 hold.

By condition 1, \(f(x)\) and \(g(x)\) are continuous on \([a,b)\). For any \(x\in(a,b)\), \(f(x)\) and \(g(x)\) are continuous on \([a,x]\) and differentiable on \((a,x)\). By condition 2, for all \(t \in (a,x)\), \(g'(t) \neq 0\). Thus, by Cauchy's mean value theorem, there exists a \(c_x \in (a,x)\) such that
\[\frac{f'(c_x)}{g'(c_x)} = \frac{f(x) - f(a)}{g(x) - g(a)} = \frac{f(x)}{g(x)}.\]
\(c_x \to a + 0\) as \(x \to a +0\) and, by condition 3, \(\lim_{x\to a+0}\frac{f'(c_x)}{g'(c_x)}\) exists. Therefore \(\lim_{x\to a+0}\frac{f(x)}{g(x)}\) also exists and
\[\lim_{x\to a+0}\frac{f(x)}{g(x)} = \lim_{x\to a+0}\frac{f'(c_x)}{g'(c_x)} =  \lim_{x\to a+0}\frac{f'(x)}{g'(x)}.\]

Next, consider the case when
  • (1') \[\lim_{x\to a+0}f(x) = \lim_{x\to a+0}g(x) = \pm \infty\]
holds instead of condition 1 above.
Let us define the constant \(L\) by
\[L = \lim_{x \to a+0}\frac{f'(x)}{g'(x)}.\]
By the definition of the right limit, for any \(\varepsilon > 0\), there exists a \(\delta_1 > 0\) such that \(a < x < a + \delta_1\) implies \(\left|\frac{f'(x)}{g'(x)} - L\right| < \varepsilon\). 
Since \(\lim_{x\to a+0}g(x) = \pm\infty\), there exists \(\delta_2 > 0\) such that \(a < x < a + \delta_2\) implies \(|g(x)| > 1\).
Let \(\delta' = \min\{\delta_1, \delta_2\}\) and \(d = a + \delta'\). By Cauchy's mean value theorem, for all \(x \in (a, d)\),  there exists a \(c_x \in (x, d)\) such that
\[\frac{f'(c_x)}{g'(c_x)} = \frac{f(d) - f(x)}{g(d) - g(x)} = \frac{\frac{f(x)}{g(x)} - \frac{f(d)}{g(x)}}{1 - \frac{g(d)}{g(x)}}.\]
By rearranging this equation, we have
\[\frac{f'(c_x)}{g'(c_x)} = \frac{f(x)}{g(x)} - r(x)\tag{eq:rx}\]
where
\[r(x) = \frac{f(d)}{g(x)} - \frac{f'(c_x)}{g'(c_x)}\cdot\frac{g(d)}{g(x)}.\]
Here, \(f(d)\) and \(g(d)\) are finite constants, and \(\lim_{x\to a+0}\frac{f'(x)}{g'(x)}\) converges to a finite value (by condition 3). Hence, by condition 1' (\(\lim_{x\to a + 0}g(x) = \pm\infty\)), \(\lim_{x\to a+0}r(x) = 0\). In other words, for any \(\varepsilon > 0\), there exists a \(\delta_3 > 0\) such that \(a < x < a + \delta_3\) implies \(|r(x)| < \varepsilon\).
Let \(\delta = \min\{\delta', \delta_3\}\). By Eq. (eq:rx), we have
\[\frac{f(x)}{g(x)} - L = \frac{f'(c_x)}{g'(c_x)} - L + r(x).\]
\(a < x < a + \delta\) implies 
\[ \begin{equation} \left|\frac{f(x)}{g(x)} - L\right| \leq \left|\frac{f'(c_x)}{g'(c_x)} - L\right| + |r(x)| < 2\varepsilon. \end{equation} \]
Therefore the right limit \(\lim_{x\to a+0}\frac{f(x)}{g(x)}\) exists and it is equal to \(L\). ■
Remark. The case of left limits \(\lim_{x\to a-0}\) can be proved similarly. Then, L'Hôpital's rule (2) should be easily proved. (Exercises) □

Corollary (L'Hôpital's rule (3))

Let \(f(x)\) and \(g(x)\) be differentiable functions on the open interval \((b, \infty)\) that satisfy the following conditions.
  1. \[\lim_{x \to\infty}f(x) = \lim_{x\to\infty}g(x) = 0.\]
  2. For all \(x > b\), \(g'(x) \neq 0\).
  3. The limit \(\lim_{x\to \infty}\frac{f'(x)}{g'(x)}\) exists.
Then, the limit \(\lim_{x\to \infty}\frac{f(x)}{g(x)}\) also exists and
\[\lim_{x\to \infty}\frac{f(x)}{g(x)} = \lim_{x\to \infty}\frac{f'(x)}{g'(x)}.\]
Furthermore, condition 1 may be replaced with
  • (1') \[\lim_{x\to \infty}f(x) = \pm\infty \text{ and } \lim_{x\to \infty}g(x) =\pm\infty\]
to have the same conclusion.

Proof. We prove the case when \(x \to \infty\) and condition 1 holds.
For the open interval \((b, \infty)\), this \(b\) can be replaced with any real number greater than \(b\). Therefore, without losing generality, we may assume \(b > 0\).

Let \(x = \frac{1}{t}\). As \(x \to \infty\), \(t\to +0\). By conditions 1 and 2,
\[\lim_{t\to +0}f\left(\frac{1}{t}\right) = \lim_{t\to +0}g\left(\frac{1}{t}\right) = 0\]
and
\[g'\left(\frac{1}{t}\right) \neq 0 ~ \text{ for all } t \in \left(0, \frac{1}{b}\right).\]
From
\[\frac{d}{dt}f\left(\frac{1}{t}\right) = -\frac{1}{t^2}f'\left(\frac{1}{t}\right)\]
and
\[\frac{d}{dt}g\left(\frac{1}{t}\right) = -\frac{1}{t^2}g'\left(\frac{1}{t}\right),\]
we have
\[\lim_{t\to +0}\frac{\frac{d}{dt}f\left(\frac{1}{t}\right)}{\frac{d}{dt}g\left(\frac{1}{t}\right)} = \lim_{t \to +0}\frac{f'\left(\frac{1}{t}\right)}{g'\left(\frac{1}{t}\right)} = \lim_{x\to \infty}\frac{f'(x)}{g'(x)}\]
so that, by condition 3, the right limit \(\lim_{t\to +0}\frac{\frac{d}{dt}f\left(\frac{1}{t}\right)}{\frac{d}{dt}g\left(\frac{1}{t}\right)}\) exists. Therefore, by L'H\^opital's rule (1), we have the limit
\[\lim_{t\to +0}\frac{f\left(\frac{1}{t}\right)}{g\left(\frac{1}{t}\right)} = \lim_{x\to\infty}\frac{f(x)}{g(x)} = \lim_{x\to \infty}\frac{f'(x)}{g'(x)}.\]
■


Comments

Post a Comment

Popular posts from this blog

Birth process

-
Image
  Defining the birth process  Consider a colony of bacteria that never dies. We study the following process known as  the birth process , also known as  the Yule process . The colony starts with \(n_0\) cells at time \(t = 0\). Assume that the probability that any individual cell divides in the time interval \((t, t + \delta t)\) is proportional to \(\delta t\) for small \(\delta t\). Further assume that each cell division is independent of others. Let \(\lambda\) be the  birth rate.  The probability of a cell division for a population of \(n\) cells during \(\delta t\) is \(\lambda n \delta t\). We assume that the probability that two or more births take place in the time interval \(\delta t\) is \(o(\delta t)\). That is, it can be ignored. Consequently, the probability that no cell divides during \(\delta t\) is \(1 - \lambda n \delta t - o(\delta t)\). Note that this process is an example of the Markov chain with states \({n_0}, {n_0 + 1}, {n_0 + 2}...
Read more

Computing integrals (3): Integration by parts

-
Image
 Sometimes, we may simplify integration by using the product rule of differentiation. This technique is called integration by parts. Theorem (Integration by parts) Let \(f(x)\) and \(g(x)\) be differentiable functions on an open interval \(I\). Then,  \(\int f(x)g'(x)dx = f(x)g(x) - \int f'(x)g(x)dx\); For any \(a, b \in I\), \[\int_a^bf(x)g'(x)dx = \left[f(x)g(x)\right]_a^b - \int_a^bf'(x)g(x)dx.\] Proof . By the product rule, \[[f(x)g(x)]' = f'(x)g(x) + f(x)g'(x)\] so \[f(x)g'(x) = [f(x)g(x)]' - f'(x)g(x).\] By integrating both sides, we have the desired results. ■ Example . Let us find \(\int x\cosh x dx\). \[ \begin{eqnarray*} \int x\cosh x dx &=& \int x(\sinh x)'dx \\ &=& x \sinh x - \int 1 \cdot \sinh x dx\\ &=& x \sinh x - \cosh x + C. \end{eqnarray*} \] Example (eg:recur) . Let us study how we can compute \[I_n = \int \frac{dx}{(x^2 + 1)^n}\] for \(n\in \mathbb{N}\). Note \[I_{n} = \int \fr...
Read more

Any covariance matrix is positive semi-definite: A proof

-
Let \(\mathbf{X} = (X_1, X_2, \cdots, X_n)^\top \in \mathbb{R}^n\) be a vector of random variables. The covariance matrix \(\Sigma\) of \(\mathbf{X}\) is a square (\(n\times n\)) matrix whose elements are covariances between the components of \(\mathbf{X}\). That is, \[\Sigma_{ij} = \mathrm{Cov}(X_i,X_j)\] where \(\mathrm{Cov}(X_i,X_j)\) is the covariance between \(X_i\) and \(X_j\) , \(i,j = 1, 2, \cdots, n\):  \[\mathrm{Cov}(X_i, X_j) = \mathbb{E}[(X_i - \mathbb{E}[X_i])(X_j - \mathbb{E}[X_j])].\] Here, \(\mathbb{E}[\cdot]\) indicates the expectation value of a random variable . Any covariance matrix has the following properties: Symmetric. That is, \[\Sigma = \Sigma^\top.\] Positive semi-definite. That is,\[\forall \mathbf{v} \in \mathbb{R}^n, \mathbf{v}^\top\Sigma\mathbf{v} \geq 0.\] See also : Positive definite matrix (Wolfram MathWorld) The symmetry is obvious from the definition of the covariance matrix.  Now, let us prove that the covariance matrix is positive semi-...
Read more