# need help on y=x^tan(x)

by Bailey
Tags: yxtanx
 P: 43 if y=x^TAN(x) is y'=TAN(x)·x^(TAN(x) - 1)·(SEC(x))^2 i got the answer by using the chain rule & the power rule by letting u=TAN(X) and y=x^u. not sure if thats the right answer since when i graph the function of y and y' , they don't show any relation with each other. i think i have to use the chain rule twice if i want to use the power rule, since x^u does not equal to ux^(u-1) (i think, not exactly sure). can someone plz help me out, thanx.
P: 265
If

$$y(x)=x^{\tan(x)}$$

then I get the following result for the derivative:

$$y'(x)=x^{\tan(x)-1}\left(x\log(x)(\sec(x))^2+\tan(x)\right)$$

or, if you prefer,

$$y'(x)=x^{\tan(x)}\left(\log(x)(\sec(x))^2+\frac{\tan(x)}{x}\right)$$

I am not quite sure what you mean with:
 think i have to use the chain rule twice if i want to use the power rule, since x^u does not equal to ux^(u-1) (i think, not exactly sure).
but if you are talking about the derivative, then the statement is true:

$$\frac{\partial}{\partial x} x^u = u x^{u-1} \frac{\partial u}{\partial x}$$.

Cheers,
Freek Suyver.
P: 321
My answer is the same as that of suyver.

 i got the answer by using the chain rule & the power rule by letting u=TAN(X) and y=x^u.
It isn't correct. You cannot let u = (tan x) and then treat it as a constant, because y is a function of x and (tan x) isn't a constant. So we shouldn't apply the power rule directly. Instead, we should first take ln(natural log) on both sides, and then do differentiation.
The first few steps should be:

$$y = x^{tan x}$$

$$ln y = (tan x)(ln x)$$ {take log on both sides}

Then you can apply the chain rule to finish the rest of the question.

P: 43

## need help on y=x^tan(x)

thanx.
P: 662
We can do it as suyver has done actually he has done the problem with partial differentiation
 It isn't correct. You cannot let u = (tan x) and then treat it as a constant
 PF Patron Sci Advisor Thanks Emeritus P: 38,429 In general, if one has y= f(x)g(x), in which both base and exponent are functions of x, one can make either of two mistakes: 1. Treat the exponent, g(x), as a constant and use the power rule y'= g(x)f(x)g(x)-1 2. Treat the base, f(x), as a constant and use the exponential rule y'= ln(f(x))f(x)g(x) The interesting thing is that the correct derivative is the sum of these two mistakes! y'= g(x)f(x)g(x)-1+ ln(f(x))f(x)g(x) as one can show by differentiating ln(y)= g(x)ln(f(x)).
 P: 28 my teacher taught me to use LN rather than LOG so i ended up with [x^tan(x)]*(tan(x)/x + sec^2(x)ln(x))
 PF Patron Sci Advisor Thanks Emeritus P: 38,429 Since there is no good reason to use logarithm to base 10 in higher mathematics (it's used in arithmetic because it works nicely with base 10 numeration), most higher mathematics texts use "log" to mean natural logarithm.
P: 28
 Originally posted by HallsofIvy Since there is no good reason to use logarithm to base 10 in higher mathematics (it's used in arithmetic because it works nicely with base 10 numeration), most higher mathematics texts use "log" to mean natural logarithm.
are you responding to me? if so, is my equation the equivalent to suyvers?
P: 265
 Originally posted by hodeez are you responding to me? if so, is my equation the equivalent to suyvers?
Yes, your equation is equivalent to mine. I always write log(x) for the e-based logarithm and 10log(x) for the 10-based logarithm.
PF Patron