#Implicit Differentiation
208 messages · Page 1 of 1 (latest)
- Ask your question and show the work you've done so far. If you've posted a screenshot of a question, specify which part you need help with.
- Wait patiently for a helper to come along.
- Once someone helps you, say thank you and close the thread with:
+close - Feel free to nominate the person for helper of the week in #helper-nominations
- Do not ping the mods, unless someone is breaking the rules.
- If you're happy with the help you got here, and the server overall, you can contribute financially as well:
it looks like to me, the best approach would be logarithmic differentiation.
y = xe^y
ln(y) = ln(xe^y)
ln(y) = ln(x) + ln(e^y)
ln(y) = ln(x) + y
then you take the derivative of both sides of the equation from here and apply chain rule
Actually, you've got it right there. Now you've just got an equation to solve.
this works even if y or x is negative due to how quotient and product rule work, some little details tho
so when we do with the respect to x do we just add x' to the x value
...what?
i don't know how to do differention to the respect to a value
"With respect to x" just means we're treating x as the variable, and everything else is either a constant or a function of x.
other way
What is the derivative of x?
1
Right. That's the derivative of x with respect to x.
oh so were only taking the derivate of x
No, we're taking the derivative of everything.
Look, do you understand what a derivative is, like, conceptually?
yea
We want to know how fast the curve is changing at a specific point.
so instanous velocity/change right
...you said you didn't understand.
yea i memorized it but it doesn't make sense
Okay, then just listen.
okay!
So we want to know how fast the curve is changing at a given point.
The problem is, the curve doesn't "change" "at a point".
Just like the phrase "instantaneous velocity" doesn't make sense, because velocity is a rate of change of position, and position doesn't change instantly.
exactly!
So we want to find the slope of a curve at a point, but we can't, because, well, we only have one point.
We can calculate the slope between two points, though.
And if both of those points are on our curve, then we call that a secant line.
Right.
So ideally, this secant line would approximate the tangent line.
And we can make it do that, by moving the second point closer to the first.
oh
as the secent line gets closer to a point it becomes the tangent line
oops
as the two points gets closer to each other it becomes tangent line
It better approximates the tangent line.
okay
What we would really like is some kind of precise, well-defined, rigorous mathematical method to get as close as we like to the tangent line. Arbitrarily close.
This is the concept we call a limit.
okay
So were you taught the limit definition of a derivative? $f'(x) = \lim_{h \to 0}\frac{f(x + h) - f(x)}{h}$?
Techie Literate
yes and no, i was thought the limit definition
which is as the x value approaches a certain number what is the y value becoming
$\lim_{x \to a}\frac{f(x) - f(a)}{x - a}$?
Techie Literate
this is the same
...as what you were taught, or as what I said before?
this equation is using two points
while the other equation is using the whole equation
Okay, I think you're trying to tell me that you were taught the definition of a limit.
yes
Which, you kind of weren't.
The real definition is more rigorous than what you were taught, but it's not strictly necessary for you to know.
I mean, that would be equal to f'(a).
okay
So the intuition here in this limit is that the point we care about is (x, f(x)), and h is the distance between that point and the second point we're using to approximate the slope.
okay
But this limit is time consuming to compute directly every time, so we just use it to prove the derivative rules and use those instead.
mm okay
Now, circling back around to your problem, the issue is that you don't actually have y as a function of x.
This equation does still describe a curve, but it's just not in a nice form that's easy to manipulate.
we do have a function y that respects x
Except we have y on both sides.
oh okay
We could technically isolate it, but that would require a non-elementary function that would just be more confusing for you.
okay
But here's the thing.
Wait, rq
Okay?
so if we have this same equation but instead of of having e^y if we had e^z could we just do the product rule to find the derivative of y
Like, z would be a third variable?
Or would z be a constant? Because then e^z is just a constant coefficient and we're looking at a line.
Oh, yeah.
so we would just do the product rule
but since y isn't a function as there is y in both sides then we have to do implicit differation
Well, y is a function, it's just not represented nicely.
And it may not be a function per se.
and thats why we have to do implicir diff
Right.
okay contiue
See, here's how equations that describe curves work; they have x and y in them, and then you plot all the points (a, b) such that when x = a and y = b, the equation is true.
So this equation does describe a curve, it just doesn't do so in standard y = f(x) form. Importantly, however, the slope of the tangent line will still be y'.
okay
So the rule is basically that we treat y as a function of x, but we don't know what function of x.
k
So you were correct that the very first step is to use the product rule, (xe^y)' = (x)' * e^y + x * (e^y)'.
okay
And that would be equal to y'.
So now what we want is to solve for y', in terms of x and y.
So what does our equation look like?
my closes thought is point slope form
No, starting from the equation given in the problem, differentiate both sides with respect to x.
That is, treat x as the variable and y as a function of x.
...show your work.
Maybe some different notation will help.
We want to take d/dx on both sides.
So what's d/dx (y)?
what does that even mean
d/dx means the derivative y to respect x
and why would we even do thatg
No, dy/dx means that.
d/dx (y) means take the derivative of y with respect to x, dy/dx is the derivative of y with respect to x.
isn't that the same thing!?
No. The first is an operation, the second is the output of the operation.
okay
If I hand you a deck of cards and tell you to deal a five card hand, my instruction to deal five cards is not equivalent to the actual five cards that end up on the table.
okay
Or better yet, just rewrite the equation with y(x) instead of y.
Oh i understand
so d/dx is like a minus sign in which you can use for a specific number/variable
and dy/dx is the derivative of the whole equation
so what do i do after the 1=e^y(x)+e^y
Nothing because you made a mistake. Go back to the previous step and rewrite y as y(x) to make it explicit that y is a function of x and see if that shakes your intuition loose.
What's the difference between f times x and f of x?
u told me to replace to y to y(x)
to make me think of something
and i did
and the only thing i could think of is product rule
...okay. y = f(x).
i agree
So replace y with f(x).
On both sides.
okay i did
$f(x) = xe^{f(x)}$
Techie Literate
what my instructor did was y'= (xe^y)' (y')
No, he didn't.
times y'
No, the entire thing isn't times y', you multiply the second term by y' because of the chain rule, because y is a function of x.
oh okay
That's the whole thing I was trying to get you to understand, that y' is not 1.
chain rule: (3a+2b)^7'= 7(3a+2b)^6 * (5)
...what?
wait im just trying to see how u use the chain rule for
give me one second
oh OKAY
Im back
so why do we use the chain rule in this scenarion
why don't we just do the product rule
Because. Y. Is. A. Function. Of. X.
No. It's because y is a function of x.
Like, okay.
I can teach you an informal way of solving this if you want
Just to solidify your intuition, so that you don't go around just assuming this by rote.
If we look at our original equation, and we pick a value for x, then that restricts y, right?
yes
So y depends on x.
yes
And this dependence is the reason we treat y as a function of x when we differentiate.
Which is why the derivative of y with respect to x is y', not 1 or anything else.
okay
Do you understand?
yes
Okay, good, I'm glad.
thanks