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!

Homework Help: Finding a parametric equation for y = f(x)

  1. Feb 15, 2004 #1
    I've searched the web for information on Parametric Equations, and most of them only give me information on how to find y = f(x) when given y = y(t) and x = x(t).

    Is there any sort of method for doing the reverse? I'm told that there are theoretically an infinite number of parametric equations for a given curve, but how do you go about finding any of them?

    One of the questions in our textbook says:

    Parametrize the curve y = f(x), x between [a,b].

    No further information was given. I think I need to know this in order to do the questions that come afterwards. For example, [tex]y^2=x^3[/tex], from (4,8) to (1,1). The answer ends up being y = (2-t)^3 and x = (2-t)^2. The solution manual gives me no explanations as to how they arrived at that answer.

  2. jcsd
  3. Feb 15, 2004 #2
    there are an infinite number of parametrizations. one easy choice is the following:

    given a curve f(x)=y

    let y=f(t) and x=t.

    this is a sort of trivial parametrization. there are lots more. what exactly are you looking for?
  4. Feb 15, 2004 #3
    I'm basically looking for the thought process behind choosing a certain type of parametrization based on a restricted domain for t. The example I gave in my question:


    The question restricted the domain of t to be between 0 and 1. And one of the answers ended up being [tex]y(t) = (2-t)^3[/tex] and [tex]x(t) = (2-t)^2[/tex].

    How did they come up with that? Is it just sort of guessing and checking?
  5. Feb 16, 2004 #4


    User Avatar
    Science Advisor
    Homework Helper

    Well I would be incliend to say:
    but that can be a bit ugly, so the people replaced by [tex]t'=t^6[/tex] so that the exponents are integers.

    With the new [tex]t[/tex] you get

    The [tex]2-[/tex] is added for the final values, but that could also be done by chaning the range for [tex]t[/tex].
  6. Feb 16, 2004 #5


    User Avatar
    Science Advisor

    NateTG is, as usual, exactly right.

    There exist many "parametrizations" for any given curve. Given that y2= x3, the simplest to find is to set
    y2= x3= t and solve for x and y separately.

    If you don't like fractions (and who does?) you might think instead
    ( )2= ( )3 and not that those will be "obviously" equal if you put t3 in the first parenthesis and t2 in the second. That is: x= t2 and y= t3 just as NateTG gave.

    That's a perfectly good parametrization- in particular, when t= 1, it gives the point (1, 1) and when t= 2, it gives the point (4,8).

    Once you have that, you could replace the "t" by any formula (in particular "2- t") and still have a parametrization for the same curve. With x= (2-t)2 and y= (2-t)3, when t= 1, we get (1,1) and when t= 0, we get (4,8), just the reverse of what we had before.
  7. Feb 16, 2004 #6
    Oooh, thank you both very much NateTG and HallsofIvy! That made perfect sense! :-)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook