Show that if ##x>1##, ##\log_e\sqrt{x^2-1}=\log_ex-\dfrac{1}{2x^2}-##

Click For Summary

Homework Help Overview

The discussion revolves around the logarithmic identity involving the expression ##\log_e\sqrt{x^2-1## for ##x>1##. Participants are exploring the relationship between this expression and ##\log_ex##, along with additional terms involving powers of ##x##.

Discussion Character

  • Exploratory, Conceptual clarification, Mathematical reasoning, Assumption checking

Approaches and Questions Raised

  • Participants discuss the manipulation of logarithmic identities and series expansions. There are attempts to express ##\log_e\sqrt{x^2-1## in terms of ##\log_ex## and other logarithmic series. Questions arise regarding the validity of certain logarithmic expansions for values of ##x## greater than 1.

Discussion Status

Some participants have provided insights into the series expansion of logarithmic functions and their limitations. There is ongoing exploration of the relationship between the left-hand side and ##\log_ex##, with various interpretations being considered. The discussion is active, with participants questioning assumptions and clarifying concepts.

Contextual Notes

There are mentions of the radius of convergence for logarithmic series, particularly regarding the expansion of ##\log_e(1+x)## and its applicability for values of ##x## greater than 1. The discussion also references the need for calculus knowledge in understanding Taylor or Maclaurin series.

RChristenk
Messages
73
Reaction score
9
Homework Statement
Show that if ##x>1##, ##\log_e\sqrt{x^2-1}=\log_ex-\dfrac{1}{2x^2}-\dfrac{1}{4x^2}-\dfrac{1}{6x^6}-\cdots##
Relevant Equations
Logarithm Rules, Logarithmic Series
##\log_e\sqrt{x^2-1}=\dfrac{1}{2}[\log_e[(x+1)(x-1)]]=\dfrac{1}{2}[\log_e(x+1)+\log_e(x-1)]##.

##\Rightarrow \log_e(x-1)=\log_e[x(1-\dfrac{1}{x})]=\log_ex+\log_e(1-\dfrac{1}{x})##

We know:

##\log_e(1+x)=x-\dfrac{x^2}{2}+\dfrac{x^3}{3}-\cdots##

##\log_e(1-x)=-x-\dfrac{x^2}{2}-\dfrac{x^3}{3}-\cdots##

Hence:

##\dfrac{1}{2}[\log_e(x+1)+\log_e(x-1)]=\dfrac{1}{2}[\log_e(1+x)+\log_ex+\log_e(1-\dfrac{1}{x})]##

##\Rightarrow \dfrac{1}{2}(x-\dfrac{x^2}{2}+\dfrac{x^3}{3}-\cdots+\log_ex-\dfrac{1}{x}-\dfrac{(\dfrac{1}{x})^2}{2}-\dfrac{(\dfrac{1}{x})^3}{3}-\cdots)##

##\Rightarrow \dfrac{1}{2}(\log_ex+x-\dfrac{x^2}{2}+\dfrac{x^3}{3}-\cdots-\dfrac{1}{x}-\dfrac{1}{2x^2}-\dfrac{1}{3x^3}-\cdots)##

This is as far as I got. I'm not sure how to perform the necessary algebra to get to

##\log_ex-\dfrac{1}{2x^2}-\dfrac{1}{4x^2}-\dfrac{1}{6x^6}-\cdots##.
 
Physics news on Phys.org
1/x^2 and higher orders of this are small for large positive x. The series is difference of LHS from log x . We know that LHS is a little smaller than log x.
 
RChristenk said:
Homework Statement: Show that if ##x>1##, ##\log_e\sqrt{x^2-1}=\log_ex-\dfrac{1}{2x^2}-\dfrac{1}{4x^2}-\dfrac{1}{6x^6}-\cdots##
Relevant Equations: Logarithm Rules, Logarithmic Series

##\log_e\sqrt{x^2-1}=\dfrac{1}{2}[\log_e[(x+1)(x-1)]]=\dfrac{1}{2}[\log_e(x+1)+\log_e(x-1)]##.
This is good until here. Note, however, that the expansion for ##\log_e(1 + x)## is only valid for ##x < 1##. You need to take a factor of ##x## out of both these logs.
 
  • Like
Likes   Reactions: MatinSAR
For large x, \ln\sqrt{x^2 - 1} \approx \ln x so start with \begin{split}<br /> \ln \sqrt{x^2 - 1} &amp;= \ln\left(x \sqrt{1 - \frac1{x^2}}\right) \\<br /> &amp;=\ln x + \frac12 \ln\left(1 - \frac 1{x^2}\right)\end{split} and since 1/x^2 &lt; 1 you can expand \ln\left(1 - \frac 1{x^2}\right) without further manipulation.
 
  • Like
Likes   Reactions: PeroK
Thanks for your replies. Why is ##\log_e(1+x)## invalid for ##x \geq 1##?
 
RChristenk said:
Thanks for your replies. Why is ##\log_e(1+x)## invalid for ##x \geq 1##?
The power series expansion is invalid for ##x > 1##, because it has a radius of convergence of ##1##.
 
Last edited:
Thread moved.
Problems involving Taylor or Maclaurin series entail knowledge of calculus.

RChristenk said:
... ##\log_e(1+x)## ...
@RChristenk, why do you write this as ##\log_e(\dots)## instead of ##\ln(\dots)##?
 
RChristenk said:
I'm looking at this free book online:

https://archive.org/details/elementaryalgebr00hall

This text was published in 1896, and use of now outdated notation is not its worst shortcoming by modern standards. The notation \ln is nowhere mentioned; \log itself is reserved for the base 10 logarithm.

Relevant to this problem, which is example 6 on page 434, in relation to the series for \ln (1 + x) the book states at para 540 on p. 433 that "except when x is vey small, [this series] is of little use for numerical calculations" and then proceeds to construct the series for \ln (N + 1) - \ln N which was examined in another thread as a method to find a series for \ln 2 (example 3 on page 434).

However the book does not state that the radius of convergence is 1. Indeed the chapter on convergence and divergence of series does not expressly discuss power series or radii of convergence as such, although the examples of the use of the ratio test[* do find radii of convergence of power series - see, eg. p. 382. So a careful reader could possibly determine for themselves that the log series does not converge for |x| &gt; 1, although there is no exercise instructing them to do so. The chapter ends with a note referring readers wanting further information on the subject to chapter XXI of the same authors' Higher Algebra.

I would suggest studying from a more modern text, which uses now-standard terminology and notation.

*The test is not expressly named as such; all of the convergence tests are named "the Nth test" in the order in which they are presented.
 
  • Like
Likes   Reactions: Mark44, RChristenk and PeroK

Similar threads

  • · Replies 6 ·
Replies
6
Views
3K
Replies
10
Views
2K
Replies
3
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
5
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 10 ·
Replies
10
Views
2K
Replies
6
Views
2K