1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Limit question (log)

  1. Aug 18, 2013 #1
    1. The problem statement, all variables and given/known data

    limx→0+ log( x - 1 +[itex]\sqrt{x^2+1}[/itex])-logx)

    Note that log is the same as loge same as ln. Couldn't find the three parallel lines symbol.


    3. The attempt at a solution

    I really don't know where to start with this one. Should I use the limit laws and bring the limit into the brackets to give

    log( limx→0+ (x) - limx→0+ 1 + √(lim x→0+x^2+1) -limx→0+ log x

    I see that this so far is most likely wrong any tips would be appreciated.
     
    Last edited: Aug 18, 2013
  2. jcsd
  3. Aug 18, 2013 #2

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    I assume you mean limx→0+ log( x - 1 +[itex]\sqrt{x^2+1}[/itex])-log(x) (i.e. the logs are not nested).
    What seems natural when faced with log(a) - log(b)?
    What can you do with [itex]\sqrt{x^2+1}[/itex] for small x?
     
  4. Aug 18, 2013 #3
    Would i make the original equationin into

    limx→0+ log ((x-1+[itex]\sqrt{x^2+1}[/itex])/x))

    = limx→0+ log( 1 - [itex]\frac{1}{x}[/itex] + [itex]\frac{\sqrt{x^2+1}}{x}[/itex])

    And then can i say that because log is continuous when x>0, that i can then put
    log(limx→0+ 1 - limx→0+[itex]\frac{1}{x}[/itex] + lim x→0+ [itex]\frac{\sqrt{x^2+1}}{x}[/itex])

    Im not sure if thats right so far, I see the inside bracket becoming ∞.
    Which means limit does not exist.
     
    Last edited: Aug 18, 2013
  5. Aug 18, 2013 #4

    HallsofIvy

    User Avatar
    Staff Emeritus
    Science Advisor

    I wouldn't. I would make it, rather
    [tex]\log\left(\frac{x+1+\sqrt{x^2+ 1}}{x}\right)[/tex]
    Now, since "logarithm" is a continuous function, it becomes a question of
    [tex]\lim_{x\to 0}\frac{x+1+\sqrt{x^2+ 1}}{x}[/tex]
    and since that "becomes "2/0" when x= 0, the limit does not exist. (More specifically, the limit is "infinity" which is just saying it does not exist in a specific way.)

     
  6. Aug 18, 2013 #5

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Shouldn't that be ... log ((x-1+...?
    You can, but doing that too soon will usually lead to indeterminate results, like +∞-∞. You need to get some cancellation done first. Can you write out an approximation for [itex]\sqrt{x^2+1}[/itex] that's valid for small x?
     
  7. Aug 18, 2013 #6
    . . . this has run-off into the ditch since early this morning and provides a nice example why it is so important to be utterly perfect in your math notation. Tell you what, no matter what you want to solve, how about we solve instead the perfectly-illustrated problem:

    [tex]\lim_{x\to 0} \log\left(\frac{x-1+\sqrt{x^2+1}}{x}\right)[/tex]

    Put the limit inside. Justify it later. Looks like the indeterminate form [itex]\frac{0}{0}[/itex]. How about L'Hospital's rule?
     
  8. Aug 18, 2013 #7
    Sorry for the headache, my second attempt above at working out had a + instead of a - after the first x in the bracket. It should read (x-1 + ......)/x

    Because the function log is continuous.

    we get

    limx→0+ ([itex]\frac{x-1+\sqrt{}x^2+1}{x}[/itex])

    = [itex]\frac{0-1+1}{0}[/itex] which is [itex]\frac{0}{0}[/itex] indeterminate form.

    Then use l'hopital rule

    limx→0+ (1+ [itex]\frac{1}{2}[/itex](x2+1)-1/2 .2x)/1

    = limx→0+ [itex]\frac{1+x}{\sqrt{}x^2+1}[/itex]

    = [itex]\frac{1}{1}[/itex]
     
  9. Aug 18, 2013 #8
    That's not entirely right you know. You forgot the log:

    [tex]\log\left(\lim_{x\to 0} \frac{1+\frac{x}{\sqrt{x^2+1}}}{1}\right)[/tex]

    And also, I am not 100% sure we can apply L'Hospital's rule for a right-handed limit. Is it continuous at zero? Maybe I'm in the ditch too. Also, the best advice I can give you this semester is to learn very good three things:

    1. Latex
    2. PF
    3. Mathematica

    I mean we need to get it right so I'll double-check my work in Mathematica. So I did, it's zero. Now, how can I prove that? And how can I write it so well so that there is no ambiguity? Latex. And how can I get excellent practice getting better? Hanging out in PF (follow rules carefully though).
     
  10. Aug 18, 2013 #9
    So if I end up with

    log(1) = 0,

    Would that mean I was right?

    or do i have to do some algebra first to solve the stuff inside the bracket?
     
  11. Aug 18, 2013 #10

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Yes, that's the right answer, but approximating √(1+x2) for small x using the usual expansion is a bit easier.
     
  12. Aug 18, 2013 #11
    would i use the l'hopital rule in my working out? and what is the usual expansion?
     
  13. Aug 18, 2013 #12

    haruspex

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Binomial.
     
  14. Aug 19, 2013 #13

    vanhees71

    User Avatar
    Science Advisor
    2016 Award

    Usually the expansion of numerator and denominator in appropriate Taylor series is simpler than using de L'Hospital's rule, which is derived from this Taylor expansions anyway. As an example take what was discussed before (and please use LaTeX to type formulae; it's so much easier to read!):
    [tex]\frac{x-1+\sqrt{x^2+1}}{x}=\frac{x-1+1+\mathcal{O}(x^2)}{x}=1+\mathcal{O}(x) \rightarrow 1 \quad \text{for} \quad x \rightarrow 0.[/tex]
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Limit question (log)
  1. IF/log question (Replies: 6)

  2. Limits of log (Replies: 1)

  3. Limit of natural log (Replies: 2)

Loading...