Lambert W function with rational polynomial

AI Thread Summary
The discussion revolves around solving the equation exp(ax+b) = (cx+d)/(ex+f) using the Lambert W function. A proposed method involves substituting u = (cx+d)/(ex+f) and manipulating the equation to express it in terms of u and v, leading to the form ve^v = (af(af+bd))/de. There is some uncertainty about the correctness of the calculations, with references to similar problems and the need for generalized Lambert functions. Additionally, links to relevant papers and resources are provided for further exploration. The conversation highlights the complexity of applying the Lambert W function to rational polynomial equations.
nlooije
Messages
1
Reaction score
0
Hi all,

During my research i ran into the following general type of equation: \exp(ax+b)=\frac{cx+d}{ex+f}
does anyone have an idea how to go about solving this equation?

thx in advance
 
Mathematics news on Phys.org
It doesn't show the steps but I got this from Wolfram
 

Attachments

  • atrttr.gif
    atrttr.gif
    2.1 KB · Views: 618
Ledsnyder said:
It doesn't show the steps but I got this from Wolfram

I think the idea is to solve for ##x## instead of ##f##.
 
Let u= \frac{cx+ d}{ex+ f}, the fraction on the right. Then, solving for x, x= \frac{d- fu}{eu- c}= -\frac{f}{e}u+ \frac{fc}{e}.

So the equation is, so far,
e^{ax+ b}= e^{-\frac{af}{e}u+ \frac{afc}{e}+ b}= u
e^{-\frac{af}{e}u}e^{\frac{afc+ bd}{d}}= u
ue^{\frac{af}{e}u}= e^{\frac{afc+ bd}{d}}

Let v= \frac{af}{e}u. Then u= \frac{e}{af}v and we have
\frac{e}{af}ve^v= e^{\frac{afc+ bd}{d}}
ve^v= \frac{af(af+ bd)}{de}

v= W(\frac{af(af+ bd)}{de}

Now work back through the substitutions to find x.
 
HallsofIvy said:
x= \frac{d- fu}{eu- c}= -\frac{f}{e}u+ \frac{fc}{e}
.

I think this calculation is wrong.
 
I have a similar problem exp(2 x)= (x+y)/(x-y) and solve for x
Generalized Lambert functions, discussed at Lambert W on wikipedia or various papers on arXiv might help, e.g.
arXiv:1408.3999v1.pdf

My notes are currently at
http://www.cwr.uwa.edu.au/~keady/Seiches/rLambert/lambertWWave.pdf
 
Last edited by a moderator:
HallsofIvy said:
Let u= \frac{cx+ d}{ex+ f}, the fraction on the right. Then, solving for x, x= \frac{d- fu}{eu- c}= -\frac{f}{e}u+ \frac{fc}{e}.

So the equation is, so far,
e^{ax+ b}= e^{-\frac{af}{e}u+ \frac{afc}{e}+ b}= u
e^{-\frac{af}{e}u}e^{\frac{afc+ bd}{d}}= u
ue^{\frac{af}{e}u}= e^{\frac{afc+ bd}{d}}

Let v= \frac{af}{e}u. Then u= \frac{e}{af}v and we have
\frac{e}{af}ve^v= e^{\frac{afc+ bd}{d}}
ve^v= \frac{af(af+ bd)}{de}

v= W(\frac{af(af+ bd)}{de}

Now work back through the substitutions to find x.
HallsofIvy said:
Let u= \frac{cx+ d}{ex+ f}, the fraction on the right. Then, solving for x, x= \frac{d- fu}{eu- c}= -\frac{f}{e}u+ \frac{fc}{e}.

So the equation is, so far,
e^{ax+ b}= e^{-\frac{af}{e}u+ \frac{afc}{e}+ b}= u
e^{-\frac{af}{e}u}e^{\frac{afc+ bd}{d}}= u
ue^{\frac{af}{e}u}= e^{\frac{afc+ bd}{d}}

Let v= \frac{af}{e}u. Then u= \frac{e}{af}v and we have
\frac{e}{af}ve^v= e^{\frac{afc+ bd}{d}}
ve^v= \frac{af(af+ bd)}{de}

v= W(\frac{af(af+ bd)}{de}

Now work back through the substitutions to find x.

May I suggest that HallsofIvy changes his name to Half-fly or Highdive, or something?
 
Keady said:
I have a similar problem exp(2 x)= (x+y)/(x-y) and solve for x
Generalized Lambert functions, discussed at Lambert W on wikipedia or various papers on arXiv might help, e.g.
arXiv:1408.3999v1.pdf

My notes are currently at
http://www.cwr.uwa.edu.au/~keady/Seiches/rLambert/lambertWWave.pdf
Added, Oct 2015: The cwr website has been taken down. The main facts are in an arXiv preprint with Istvan Mezo
 
Last edited by a moderator:
Back
Top