By using inverse trig identities, it can be shown that $sec(x)*cos(x)=1$. However, when $x=π/2$, the resulting function would be $$(1/0)*(0/1)=1,$$ and cross multiplying yields $$1*0=1.$$ I'm not sure where I went wrong in this example but there doesn't seem to be any apparent gaps in the graph of $sec(x)*cos(x)$. What am I doing wrong?

  • 9
    $\begingroup$ $\sec(\pi/2)$ is not defined. $\endgroup$ – sinbadh Jan 4 '16 at 16:42
  • 2
    $\begingroup$ The problem is that it is division by zero. You can not divide by zero and expect good things to happen. $\endgroup$ – JMoravitz Jan 4 '16 at 16:45
  • 1
    $\begingroup$ It cannot be defined as $\frac10$, since that's not a valid fraction. $\endgroup$ – Arthur Jan 4 '16 at 16:45
  • 2
    $\begingroup$ Why is this getting downvoted? Though naïve, this is a perfectly valid question, and the OP has clearly demonstrated that he/she has put thought into it. $\endgroup$ – wgrenard Jan 4 '16 at 17:43
  • 1
    $\begingroup$ It shouldn't be downvoted and we shouldn't insult naive people. I probably shouldn't yell at them either. But I do get emphatic about basic points. $\endgroup$ – fleablood Jan 4 '16 at 17:48
  1. The domain of $\sec x$ does not include $\frac{\pi}{2}$; $\sec\frac{\pi}{2}$ is not defined. More information here.

  2. Even if you didn't know that, or forgot it, $\frac{1}{0}$ is not defined.

  3. Even if you forgot those things, $0\cdot\frac{1}{0} = \frac{0}{0}$ by your logic, which is undefined.


The terms $1/0,0/0,\infty$ are indeterminate or undefined and cant be used for algebra eg $6\times 0=5\times 0$ cancelling zero we get $6=5$ so you get why they cant be used !

  • $\begingroup$ @cr3 i have edited $\endgroup$ – Archis Welankar Jan 4 '16 at 16:59
  • $\begingroup$ @ArchisWelankar Its nice . $\endgroup$ – Tesla Jan 4 '16 at 16:59
  • $\begingroup$ $\infty$ by itself is not indeterminate nor undefined, but you're right that it can't be used for algebra because it's not a number. $\endgroup$ – cr3 Jan 4 '16 at 17:01
  • $\begingroup$ Thanks @ Tesla and yes that was my motive algebraically useless numbers @ cr3 $\endgroup$ – Archis Welankar Jan 4 '16 at 17:06
  • 1
    $\begingroup$ You could just say "Undefined terms like $1/0$, indeterminate terms like $0/0$ and not actual numbers like $\infty$ can't be used..." rather than mislabel $\infty$ as indeterminate or undefined. $\endgroup$ – cr3 Jan 4 '16 at 17:10

$m$ divided by $n$ is the specific number, $q$, such that $qn = m$. So $m$ divided by $0$ is the specific number, $q$, such that $q*0 = m$. There is no such number so division by 0 is impossible and 1/0 is undefined.

Now, trigonometry. You have a unit circle. You have an angle $\theta$. The points on the unit circle corresponding to the angle $\theta$ are $(x, y) = (\cos \theta, \sin \theta)$ and there is an inscribed right triangle corresponding to that angle with sides $x = \cos \theta$, $y = \sin \theta$ and hypotenuse = 1. That's basic.

So... Draw a tangent line from the (x,y) point that is tangent to the circle. Extend that line until it crosses the x-axis (!!Warning!! !!Warning!! Who ever told you it would cross the x-axis?) (Oh, shut up, voice in my head; most lines will cross the x-axis.)... extend that line until it crosses the x-axis. This forms another right angle that is similar to the first. The tangent line from $(x,y)$ to the x-axis coresponds to the $x = \cos \theta$ side of the first triangle. The radius of the unit circle (from the origin to $(x,y)$) on the new triangle corresponds with the $y = \sin \theta$ side of the original triangle. The x-axis from the origin to where the line intersects the x-axis corresponds to the hypotenuse of the original triangle.

Doing a bit of geometry with realize the scaling factor of this new triangle is $1/\cos \theta$. So the distance of the tangent point to the x-axis is $\sin \theta /\cos \theta$. We call that $\tan \theta$ (because it's the length of a tangent chord!!!). And the length from the $(0,0)$ origin to where the tangent line intersects the x-axis is $1/\cos \theta$. We call that $\sec \theta$ (because it is the length of a secant chord !... wait, what the heck is a "secant chord"? .. ).

Let me repeat that as I'm sure your eyes are glazing over.

$\sec \theta$ is the distance from the origin to where the tangent line intersects the x-axis.

Okay, so let $\theta = \pi/2$. $\sec \pi/2$ is the distance from the origin to where the tangent line intersects the x-axis. But the tangent line is parallel to the x-axis!!!! The tangent line never intersects the x-axis and there is no such distance!

(Maybe I should have listened to that voice in my head. Wouldn't be the first time. Of course, there was the time when it told me to... well, never mind.)

So $\sec \pi/2$ is undefined.

Just like $1/0$ is undefined.


There are many similarities. $\sec \theta = 1/\cos \theta$ so $\sec \pi/2$ would $= 1/0$ if such a thing were possible.

Also notice as $\theta$ gets closer and closer to $\pi/2$ the tangent line gets larger and larger and $\sec \theta$ gets larger and larger and approaches infinity.

Likewise if you are dividing $m$ by $n$ and $n$ gets smaller and smaller and closer to 0 then $m/n$ gets larger and approaches infinity.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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