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

Finding bend(maxima/minima) in plotted curve

  1. Jul 27, 2007 #1
    I want to write a computer program or subroutine to find the number of bends or maxima/minima on a plotted curve of points. How can I use the points defining the curve to do this. I would expect that the change in slope would tell me, but how significant of a change should I be looking for? I think this is the right approach. Any help would be appreciated.
     
  2. jcsd
  3. Jul 27, 2007 #2

    Kurdt

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Well when the slope changes sign is how we usually tell there is a local maximum or minimum.
     
  4. Jul 27, 2007 #3
    Well, what if the curve does not reverse directions, but only has a steeper slope?
     
  5. Jul 27, 2007 #4

    Kurdt

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Then the slope will not have changed sign. I'm not sure what you mean by a curve reversing direction. Could you clarify?
     
  6. Jul 27, 2007 #5
    Well, I suppose in a sine wave the curve oscillates and reverses direction.
     
  7. Jul 27, 2007 #6

    Kurdt

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Ok well a sine wave consists of a lot of local maxima and minima. at these maximum and minimum points the gradient of the curve changes sign. That would indicate the curve reversing direction as you put it. If the curve doesn't reverse direction but the gradient gets steeper then there is no local maximum or minimum. Perhaps you are talking of a verticle asymptote however?
     
  8. Jul 27, 2007 #7
    Perhaps it is a vertical asymptote. Actually it is a pump curve. I suppose this would simply be a 1st order curve or curve of one degree still?

    OK, yes, I suppose a pump curve would be a vertical asymptote. So how would I find the order of the equation based on the plot. Should I assume it to be a second order equation?
     
    Last edited: Jul 27, 2007
  9. Jul 28, 2007 #8

    Kurdt

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I think you'll need an engineer to help you with that. I'm not sure what a pump curve is. From having a quick google it doesn't look as if they have any verticle asymptotes.
     
  10. Jul 28, 2007 #9
    OK, it looks like it will be a 1st order or degree equation. Maybe it will curve or maybe it will be straight, but it probably should not have more than one degree or order.
     
  11. Jul 29, 2007 #10

    Kurdt

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    It might help if you could posta picture of the curve you are trying to describe.
     
  12. Jul 29, 2007 #11
    Yes, I was thinking of posting some sample curves, but I could not figure out how to do it. How do I upload a file here? Would pasting it work?
     
  13. Jul 29, 2007 #12

    Kurdt

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    You can host an image on an image hosting site and use image tags, or there is a facility to upload files but their size is limited and they need to be approved by a mentor before anyone can see them which sometimes takes a while.
     
  14. Jul 29, 2007 #13
  15. Jul 29, 2007 #14

    cristo

    User Avatar
    Staff Emeritus
    Science Advisor

  16. Jul 29, 2007 #15
    So how can I determine the degree of the curve equations and better yet solve for the equations of the curves?
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Finding bend(maxima/minima) in plotted curve
  1. Maxima-minima problems (Replies: 5)

  2. Maxima minima (Replies: 3)

  3. Global maxima & minima (Replies: 5)

Loading...