Sean Fitzpatrick |
---|
University of Lethbridge |
![]() |
Identify the curve defined by the parametric equations:
\(x = 2t-1, y=-t+4, t\in \R\)
\(x=\cos(2t), y=\sin(2t), t\in [0,\pi]\)
Defined as the locus (set) of points equidistant from a line (the directrix) and a point (the focus).
If the vertex is at \((0,0)\text{,}\) the focus is at \((0,p)\text{,}\) and the directrix is \(y=-p\text{,}\) then the parabola has equation
An ellipse is the set of all points \(P\) such that the sum of the distances from \(P\) to two points \(F_1,F_2\text{,}\) called the foci is constant.
If the centre is at \((0,0)\text{,}\) and the foci are at \((\pm c, 0)\text{,}\) the ellipse has equation
Vertices are at \((\pm a, 0)\) and \((0, \pm b)\text{.}\)
Sketch the ellipse \(\dfrac{x^2}{9}+\dfrac{y^2}{4}=1\text{.}\)
A hyperbola also has two foci. This time, we look for the set of all points where the difference of the distances between each point and the foci is constant.
Result:
Sketch the hyperbola \(\dfrac{x^2}{9}-\dfrac{y^2}{4}=1\text{.}\)
Usually we describe curves “explicitly”, as graphs of functions: \(y=f(X)\text{.}\)
This is rather restrictive: lots of intesting curves are not graphs! (Like circles, ellipses, hyperbolas...)
One option is defining curves implicitly using equations of the form \(f(x,y)=c\text{.}\)
Another is to define them parametrically: we define \(x\) and \(y\) as functions of a third variable \(t\text{.}\)
An example you've seen: \(x=\cos\theta\text{,}\) \(y=\sin\theta\text{.}\)
From linear algebra: \(\langle x,y\rangle = \langle 2,3\rangle + t\langle 5,1\rangle\)
\(x=t^2-t, y=1-t^2, t\in [-3,3]\)
\(x=t^3-t+3, y=t^2+1, t\in [-2,2]\)
\(x=\cos(t), y=\sin(2t), t\in [0,\pi]\)
Sometimes we can gain insight on a parametric curve by converting back to an equation relating \(x\) and \(y\text{.}\)
Examples:
\(x = 3\cos(t)-1, y=2\sin(t)+3, t\in [0,2\pi]\text{.}\)
\(x=\cos(t), y=\cos(2t), t\in [0,\pi]\)
\(x=\dfrac{1}{2+t}, y = \dfrac{3t+9}{3+t}\)
Note: plotting curves on a computer often relies on being able to parametrize them. But (except for graphs) this isn't always easy!
A curve \(x=f(t), y=g(t)\) is considered smooth if \(f\) and \(g\) are both differentiable, and \(f'(t)\) and \(g'(t)\) are never simultaneously zero.
Some people will also require that a smooth curve has no self-intersections. Finding points of self-intersection is an algebraic nightmare.
Find the points where \(x = t^2-4t, y=t^3-2t^2-4t\) is not smooth.
Find the points where \(x = \cos(t), y=2\cos(t)\) is not smooth.
Our definition of smooth curve is natural in one sense: it means we can find tangent lines!
Slope of the tangent is given by \(\dfrac{dy}{dx}\text{.}\) Chain rule gives:
If either \(x'(t)\) or \(y'(t)\) is undefined, we can't compute the slope of the tangent. If both are zero, the slope is indeterminate.
(If \(x'(t)=0\) but \(y'(t)\neq 0\text{,}\) we have a vertical tangent!)
(Aside:) a vector in the direction of the tangent line at \((x(t_0),y(t_0))\) is \(\langle x'(t_0), y'(t_0)\rangle\text{.}\)
Find the equations of the tangent lines as indicated:
\(x=t^2-1, y=t^3-t\text{,}\) at \(t=0\) and \(t=1\)
\(x=\tan(t), y=\sec(t)\text{,}\) at \(t=\pi/4\text{.}\)
We can compute area as usual. For area under a curve: \(A = \int_a^b f(x)\,dx\text{.}\)
If \(x=g(t), y=f(g(t))\text{,}\) get \(A = \int_{t_0}^{t_1} f(g(t))g'(t)\,dt\text{.}\)
This still makes sense for a general parametric curve if we're careful about interpretation.
For a closed curve \(x=g(t), y=h(t)\) with counterclockwise orientation, the area enclosed is
Find the area encolosed by the ellipse \(\di\frac{x^2}{a^2}+\frac{y^2}{b^2}=1\text{.}\)
Find the area encolosed by the “teardrop” in the curve \(x=t^3-t, y=t^2-1\text{.}\)