Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I numerically solved the transcendental equation $$\ln x-\sqrt{x-1}+1=0$$ and obtained an approximate value of its positive real root $$x \approx 14.498719188878466465738532142574796767250306535...$$

I wonder if it is possible to express the exact solution in terms of known mathematical constants and elementary or special functions (I am especially interested in those implemented in Mathematica)?

share|improve this question
    
Try Lambert W function. –  ᴊ ᴀ s ᴏ ɴ Jun 24 '13 at 6:11
    
The Inverse Symbolic Calculator, to my surprise, did not recognize it. –  André Nicolas Jun 24 '13 at 6:16
    
Mathematica couldn't solve it with Solve[] and Reduce[]. –  Fixed Point Jun 24 '13 at 8:21
    
@ᴊᴀsᴏɴ Have you been able to express it in terms of Lambert W? –  Vladimir Reshetnikov Jun 25 '13 at 21:27
    
@AndréNicolas Does "to my surpise" mean that you were able to find a rather simple form for the root? –  Vladimir Reshetnikov Jun 25 '13 at 21:29
show 2 more comments

1 Answer 1

up vote 17 down vote accepted
+200

Yes, it is possible to express this root in terms of special functions implemented in Mathematica.


Start with your equation $$\ln x-\sqrt{x-1}+1=0,\tag1$$ then take exponents of both sides $$x\ e^{1-\sqrt{x-1}}=1.\tag2$$ Change the variable $$z=\sqrt{x-1}-1,\tag3$$ then plug this into $(2)$ and divide both sides by $2$ $$\left(\frac{z^2}2+z+1\right)e^{-z}=\frac12.\tag4$$ Now the left-hand side looks very familiar. Indeed, as it can be seen from DLMF 8.4.8 or the formulae $(2),(3)$ on this MathWorld page, it is a special case (for $a=3$) of the regularized gamma function $$Q(a,z)=\frac{\Gamma(a,z)}{\Gamma(a)},\tag5$$ implemented in Mathematica as GammaRegularized[a, z].

Its inverse with respect to $z$ is denoted as $Q^{-1}(a,s)$ and implemented in Mathematica as InverseGammaRegularized[a, s]. We can use this function to express the positive real root of the equation $(4)$ is a closed form $$z=Q^{-1}\left(3,\ \frac12\right).\tag6$$

Finally, using $(3)$ we can express the positive real root of your equation $(1)$ as follows: $$x=\left(Q^{-1}\left(3,\ \frac12\right)+1\right)^2+1.\tag7$$

The corresponding Mathematica expression is

(InverseGammaRegularized[3, 1/2] + 1)^2 + 1

We can numerically check that substitution of this expression into the left-hand side of the equation $(1)$ indeed yields $0$.

I was not able to express the result in terms of simpler functions (like Lambert W-function).

share|improve this answer
4  
A heuristic check confirming that the expression we found is indeed the root of the equation can be done in Mathematica using PossibleZeroQ[With[{x = (InverseGammaRegularized[3, 1/2] + 1)^2 + 1}, Log[x] - Sqrt[x - 1] + 1]] (* True *). –  Vladimir Reshetnikov Jun 25 '13 at 22:19
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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