Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Functions and their Extrema

  1. Dec 10, 2011 #1
    Hey guys,

    I have learned that for a function f defined on an open interval I(a, b), an absolute extremum may not occur. Is this because the domain may take on every value greater than a or less than b without ever equaling a or b?

    [EDIT]

    Disregard the question above. I just realized. For a function such as f(x) = x2, there is no absolute maximum because as x increases, f(x) increases indefinitely.
     
    Last edited: Dec 10, 2011
  2. jcsd
  3. Dec 10, 2011 #2

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Check your ideas:
    Does [tex]e^{-x^2}+x^2[/tex] ... have an absolute maximum at x=0?
     
  4. Dec 10, 2011 #3
    Wouldn't that be a minimum?
     
  5. Dec 10, 2011 #4

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    how do you figure?
     
  6. Dec 10, 2011 #5
    At 0, y=1 and for all other x less than or greater than zero, y is equal to or greater than 1. Correct?
     
  7. Dec 11, 2011 #6

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    It's pretty flat about x=0 isn't it? ... how about:
    [tex]2e^{-x^2} + \frac{1}{2}x^2[/tex]
     
  8. Dec 11, 2011 #7
    For this, wouldn't the absolute minima be at the troughs around x= 1 and x= -1 and the absolute maximum doesn't exist?
     
  9. Dec 11, 2011 #8

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    There is a local maximum at x=0, but it ain't global.

    If there is one "absolute" or global minima, then it is "unique"... here there is no unique global minimum. Well done.

    Technically they are at: [itex] x = \pm \sqrt{-\ln{(1/4)}} \approx \pm 1.1773[/itex] but I'm not really looking for exact here, just understanding.

    There is no new math here, it's just a matter of getting used to the common language.
     
  10. Dec 11, 2011 #9
    How did you post the mathematics in the format that you did? Tags?
     
  11. Dec 11, 2011 #10

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    If you quote my message you'll see the tags.

    The bb code is [tex] (pronounced tech - as in technology) - to put equations inline like I did use [itex]. Everything written inside those tags is LaTeX (lay-tech).
    This is an academic standard way to mark-up documents for publication: really worth learning. The forum server has an engine which can turn LaTeX markup into a graphic... It makes writing out math very very simple and clear.

    There are loads of tutorials etc around the web, and you can get a "Latex" program for any platform for you to use at home.
    The only problem is, when you tell someone "I am really into latex." they look at you funny.
     
  12. Dec 11, 2011 #11
    OK, so then allow me to do another example.

    for [tex]2e^{-x^2} + \frac{1}{2}x[/tex],

    there wouldn't be any global extrema. Zero would be a local maximum, and the the local minimum would depend on the open interval. Is this correct?
     
  13. Dec 11, 2011 #12

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    Extrema are all local to themselves, so it is a fair characterization without specifying the region. I checked and it turns out that you can have more than one global extrema ... which is why the term "unique" gets applied when there is only one.

    In [itex]f(x)=x(x-1)(x+1)[/itex] there are two extrema, neither are global.

    The fun thing is that the function can be anything ... eg. does the dirac delta function have any extrema?
     
  14. Dec 11, 2011 #13
    So you mean that all extrema are local, but not all local extrema are global?
     
  15. Dec 11, 2011 #14

    Simon Bridge

    User Avatar
    Science Advisor
    Homework Helper
    Gold Member
    2016 Award

    That would be right.

    I use local and global for relative and absolute - I don't think it matters.
    But it is useful to play around and explore to deepen your understanding.
    It's usually more important to be able to express words in math than math in words though.

    So now you can use the language, and you learned to use LaTeX.
    That should be enough for now :)
     
  16. Dec 11, 2011 #15
    I agree. I am better at understanding the concept and creating abstractions in my mind than I am at mathematical terminology.

    Thank you for all of your help.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Functions and their Extrema
  1. Extrema of function (Replies: 1)

  2. Extrema of functionals (Replies: 5)

Loading...