0

For a function

$$ f(x)=x-6\sqrt{x-1}, $$

can $x=1$ be considered a critical number, and is $f(1)$ a local maximum?

The definition my textbook gives me of a critical number is a number that is an interior point of the domain of $f$ where $f'$ is either zero or undefined. If $f$ is continuous in $[1,\infty)$, can $x=1$ be considered as an interior point?

rayank97
  • 71
  • 6
  • 2
    Well, sometimes those introductory book on Calculus make up definitions slightly different from one another, sometimes inconsistent with themselves. So, you would need to check. Can you quote exactly the definitions of "critical number", "local maximum", and "interior point"? Certainly $f(1)\geq f(x)$ for all $x\in[1,+\infty)$ that are close enough to $1$. If they defined "local maximum" in this spirit it should be. – plop Oct 03 '22 at 21:17
  • 2
    I am used to call a point interior of a set $A$ as a subset of a larger set $X$ that has a notion of open subsets. I would call $1$ a point that is interior of $[1,\infty)$ as a subset of $[1,\infty)$. But also $1$ is not interior of $[1,\infty)$ as a subset of $\mathbb{R}$. – plop Oct 03 '22 at 21:19

1 Answers1

1

First, a definition:

Definition: A point $x$ is an interior point of a set $E \subseteq \mathbb{R}$ if there exists some $r > 0$ such that $$ B(x,r) = \{ y : |x-y| < r \} \subseteq E. $$

That is, a point is in the interior of $E$ if that point is contained in an open ball which is entirely contained within $E$.

The natural domain of $f$, defined by the formula $$ f(x) = x - 6\sqrt{x-1},$$ is the interval $[1,\infty)$ (this is the largest set of real numbers such that the formula defining $f$ gives a real number). Note that any ball in $\mathbb{R}$ which contains $x = 1$ must also contain some point $y=1-\varepsilon$, where $\varepsilon > 0$. Hence there is no $r > 0$ such that $$ B(1,r) \subseteq [1,\infty). $$ As such, $x=1$ is not an interior point of the domain of $f$. Therefore, by the definition provided, $x=1$ is not a critical point of $f$.

However, that does not mean that $f(1)$ is not a (local) maximum of $f$.

Definition: Let $f : D \to \mathbb{R}$ be a function (that is, $f$ is a real-valued function with domain $D$). A real number $M$ is a local maximum of $f$ if

  1. there is some $c \in D$ such that $f(c) = M$, and
  2. there is some $r > 0$ such that $f(c) \ge f(x)$ for all $x \in D$ with $|x-c| < r$.

It can be shown directly that there is an interval $[1, b)$ on which $f(x) \le 1$, which is sufficient to show that $f$ has a local maximum at $x=1$. It is instructive to work this this out from the definition, rather than appealing to theorems.

Again, the goal is to show that $f(x) \le 1$ on some (open) interval containing $1$. Doing some basic algebra, \begin{align} f(x) \le 1 &\iff x - 6\sqrt{x-1} \le 1 \\ &\iff (x-1) - 6\sqrt{x-1} \le 0 \\ &\iff \sqrt{x-1} \left( \sqrt{x-1} - 6 \right). \end{align} Observe that $\sqrt{x-1} \ge 0$ for all $x \ge 1$, and that $$ \sqrt{x-1} - 6 < 0 \iff \sqrt{x-1} < 6 \iff x-1 < 36 \iff x < 37. $$ From this, it follows that $f(x) \le 1$ on the interval $[1,37)$. Therefore $f(1) = 1$ is a local maximum of $f$.

Note that this compares favorably to a GeoGebra sketch of the problem:

enter image description here

Note, also, that one can appeal to theorems. Specifically, it can be shown that if $f$ is differentiable on an interval $(a,b)$ and $f'(x) < 0$ on that interval, then $f$ is decreasing. A quick computation shows that $f'(x) < 0$ on the interval $(1,10)$, hence $f$ is decreasing on that interval. It can then be inferred that $f(1)$ is a local maximum of $f$. However, there are several "gaps" in this argument, which do need to be filled in.

  • Following this explanation, I'm faced with a conundrum. If $x=1$ is not an interior point, then it cannot be a critical point. And local extrema can only occur at critical points, or am I mistaken? And if so, how do we justify $f(1)$ being a local maximum of the function?

    Following @user85667 's comment, I checked for the definition of a local maximum in the textbook. It is as follows:

    "A function $f$ has a local maximum value at a point $c$ within its domain $D$ if $f(x)\le f(c)$ for all $x\in D$ lying in some open interval containing $c$."

    – rayank97 Oct 04 '22 at 07:39
  • According to this definition, your explanation of $f(1)$ being a local maximum holds, but then I am faced with the conundrum once again when I move forward to reach the definition of a critical point:

    "An interior point of the domain of a function $f$ where $f'$ is zero or undefined is a critical point of $f$."

    However, this is preceded by a short explanation where the textbook states:

    "... the only places where a function $f$ can possibly have extrema (local/global) are:

    1. Interior points where $f'=0$
    2. Interior points where $f'$ is undefined
    3. endpoints of the domain of $f$."
    – rayank97 Oct 04 '22 at 07:45
  • The textbook claims that this definition summarizes results 1, 2, and 3. Then, does this mean the textbook considers $1$ to be an interior point/critical number? Because that's the only way I could consider $1$ to be a critical number, and $f(1)$ to be a local maximum. – rayank97 Oct 04 '22 at 07:49
  • Who says that minima and maxima can only occur at critical points? Fermat's theorem on stationary points asserts that if $f$ is differentiable and $f$ has a local extremum at $c$, then $f'(c) = 0$. You are asserting the converse, which is false. – Xander Henderson Oct 04 '22 at 12:16
  • Look carefully at your book, again. Extrema can happen in one of three places: (1) interior points of the domain where the derivative is zero; (2) interior points of the domain where the derivative fails to exist; or (3) boundary points. Your book's definitions are consistent---boundary points are not critical points, but they are places where you might find extrema. – Xander Henderson Oct 04 '22 at 12:19