Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

the following is a question from my textbook on vectors:

enter image description here

EDIT: Added text, so that the post is self-contained even without the picture.

The points $A$ and $B$ have position vectors $\begin{pmatrix}2\\9\\t\end{pmatrix}$ and $\begin{pmatrix}2t\\5\\3t\end{pmatrix}$ respectively.
a. Find $\vec{AB}$.
b. Find, in terms of $t$, $|\vec{AB}|$.
c. Find the value of $t$ which makes $|\vec{AB}|$ a minimum.
d. Find the minimum value of $|\vec{AB}|$.

My issue is with part C. I accept the textbook's solution, as mine gives an incorrect answer but do not understand why their method is acceptable. Since the exercise is to find the value of t that makes vector AB a minimum, why is it acceptable to square vector AB and then differentiate, rather than just differentiating vector AB as it is, using the chain rule? I see how this leads to the problem of two t values, rather than one, but I didn't feel that it would be acceptable to just ignore one by squaring, and changing the original formula. What am I misunderstanding here?

enter image description here

EDIT: Added text.

(a) $\vec{AB}=\mathbf{b}-\mathbf{a}=\begin{pmatrix}2\\9\\t\end{pmatrix}-\begin{pmatrix}2t\\5\\3t\end{pmatrix} =\begin{pmatrix}2-2t\\-4\\2t\end{pmatrix}$

(b) $\begin{align}|\vec{AB}|&=\sqrt{(2-2t)^2+(-4)^2+{2t}^2}\\ &=\sqrt{4t^2-8t+4+16+4t^2}\\ &=\sqrt{8t^2-8t+20} \end{align}$

(c) Let $|\vec{AB}|^2=p$, then $p=8t^2-8t+20$ and $\frac{\mathrm{d}p}{\mathrm{d}t}=16t-8$


share|cite|improve this question
up vote 6 down vote accepted

It's acceptable to square and then differentiate because squaring is a monotone function on positive arguments (that is, $0\lt a\lt b$ if and only if $0\lt a^2\lt b^2$), so a value of $t$ maximizes the original if and only if it maximizes the square.

It's good to square because it's easier to differentiate the square than it is to differentiate the original.

share|cite|improve this answer

It is true that squaring is easier. But one does not have to do it (I don't see how you get two solutions if you do not square it). Having $$|\vec{AB}|= \sqrt{8 t^2 -8 t +20},$$ we obtain using the chain rule $$\frac{d}{dt} |\vec{AB}| = \frac{16 t -8}{2\sqrt{8 t^2 -8 t +20}}.$$ The denominator is always positive, as $8 t^2 -8 t +20$ assumes its minimum $18$ for $t=\frac{1}{2}$. The derivative is therefore only zero for $$16 t = 8$$ which is the same condition you will obtain by first squaring the equation.

share|cite|improve this answer

The OP is, of course, bound by what his textbook tells him to do. However, it also looks like part b of the problem is just leading up to part c, which is what whoever wrote the textbook really wanted to ask. IMHO it is more interesting to solve this type of a problem with algebra/geometry alone, i.e. without resorting to methods from calculus such as differentiation. If we can, that is :-)

Below I will switch to row vector notation in the interest of saving space (not to mention making the typesetting easier).

After solving part a, we want to find a value of the variable $t$ that minimizes the length of the vector $r=(x,y,z)=(2t-2,-4,2t)=(-2,-4,0)+t(2,0,2)$. When we vary $t$ these points move along the line $L$ going thru the point $(-2,-4,0)$ and in the direction given by the vector $s=(2,0,2)$. We are to find the point $r$ of $L$ that is as close to the origin as possible. If you draw a picture, it is clear that minimal distance occurs, when $r$ is perpendicular to the vector $s$. Thus we get the equation $$ r\cdot s=0\Leftrightarrow 2(2t-2)+2(2t)=0\Leftrightarrow 8t-4=0, $$ as two (non-zero) vectors are perpendicular to each other, if and only if their dot product vanishes. From this we can solve $t=2$, and proceed as above.

Note: Also the derivative of $p=r\cdot r$ with respect to the parameter $t$ equals $$dp/dt=r\cdot\frac{dr}{dt}+\frac{dr}{dt}\cdot r=2s\cdot r,$$ because $dr/dt=s$. Therefore the two solutions are equivalent.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.