Find X in Y=X+e^X: Log It & Solve

  • Thread starter Thread starter Gelsamel Epsilon
  • Start date Start date
AI Thread Summary
To find X in the equation Y = X + e^X, taking the logarithm leads to an incorrect assumption that Log Y = Log X + X. The correct approach involves recognizing that X can be implicitly defined as a function of Y, given that Y(X) is bijective. Differentiating both sides yields a differential equation for X(Y), specifically dX/dY = 1/(1 + e^(X(Y))). This allows for the formulation of an integral equation, X(Y) = ∫(1/(1 + e^(X(Y)))) dY, which can provide approximations for X near specific values of Y. The discussion also touches on the Lambert W function as a potential solution method.
Gelsamel Epsilon
Messages
315
Reaction score
0
How do I find X with respect to Y if;

Y = X + e^X?

If I log it then I get Log Y = Log X + X, if this is the correct step how do I isolise X from here?
 
Mathematics news on Phys.org
Taking the log of the right-hand side is not log x + x, but instead log(x + e^x).
 
****, stupid mistake of mine :-).

I still can't see that helping me at all though :-(.
 
It's not always possible to explicitely invert a function.
 
Last edited:
So I can implicitly get one?

y' = 1+e^x
y'' = e^x

y - y'' = x?
 
The equation that implicitly defines X as a function of Y is the original equation. Y = X + e^X. The fact that Y(X) is in bijection with its image assures you that X(Y) exists.

It's not the case with say Y(X)=X² because X(Y) would be 2-valued: ±\sqrt{Y}.

But your idea with differentiating bth sides was good. However, I think it's more interesting to acknowledge X(Y) exists, and then differentiate the equation wrt Y! I get

1=\frac{dX}{dY}+\frac{dX}{dY}e^{X(Y)}

\frac{dX}{dY}=\frac{1}{1+e^{X(Y)}}

This way you have a differential equation for X(Y), or inversely, an integral equation.

X(Y)=\int\frac{1}{1+e^{X(Y)}}dY

I don't know if something can be done with those, but there's probably a way to at least extract some information or approximation. For instance, we know from Y(X) that when Y is around 1, Y is small, so you can keep only the first few terms in the series of exp(X(Y)) and integrate to get an approximation of the form of X(Y) in the nbhd of 1.
 
Gelsamel Epsilon said:
How do I find X with respect to Y if

what on Earth does that even mean?
 
I hope this doesn't get into semantics.
 
May we know where this problem of yours comes from?P.S.This is nonsense, don't pay attention:

quasar987 said:
For instance, we know from Y(X) that when Y is around 1, Y is small, so you can keep only the first few terms in the series of exp(X(Y)) and integrate to get an approximation of the form of X(Y) in the nbhd of 1.
 
  • #10
Got to choose from 5 different questions to do animation type exercises in Maple. Chose number 5 but we had problems restricting the domain in the 2nd function to the time in the first function. The first function was in the form of y = t +e^t. We decided it was too hard and started on a different, easier, question. But it still bugged me that we couldn't get a function for t(y). Obviously we weren't supposed to do it like that, good idea we chose an easier one then eh? :-).
 
  • #11
y=x + e^x
dy/dx=1+e^x
dx/dy = 1/(1+e^x)
 
  • #12
1. Subt z=e^x to get : y = z + ln(z)

2. Exponential each side to get : e^y = z e^z

3. From the definition of the LambertW function : z = LambertW( e^y)

4. So x = ln( LambertW( e^y ) )
 
Last edited:

Similar threads

Replies
3
Views
2K
Replies
13
Views
2K
Replies
3
Views
1K
Replies
44
Views
4K
Replies
2
Views
13K
Replies
1
Views
985
Replies
6
Views
1K
Back
Top