Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon. Entire thread

Graph problem

Name: Anonymous 2007-12-03 22:54

I'm having a little problem with the equation for a graph. I'm looking for something where the line is completely horizontal at point 1,1 and completely vertical at 0,0. So far I have y=-x^2+2x which gives a horizontal line at 1,1, but not a vertical line at point 0,0 - I don't think it can be done with a parabola, but I have no idea where to go from here. Pardon my noobishness. Some help please?

Name: Anonymous 2007-12-03 23:32

So for a line (by which I assume you mean tangent) to be completely vertical means that the limit of derivative must be ∞, so the easiest graph I can think of which satisfies this is a circle of radius 1 centered at (1,0) I belive that is (x-1)^2+y^2=1.

Name: Anonymous 2007-12-03 23:35

It can't be done with a parabola (or any polynomial for that matter), since their derivatives are always defined everywhere, making a vertical tangent impossible.

f(x) = x^(1/3) + x/3 satisfies the requirement.

Name: Anonymous 2007-12-03 23:55

You win multiple internets.

Name: Anonymous 2007-12-04 9:23

>>3
That doesn't even go through (1,1). Circle works though.

Name: Anonymous 2007-12-04 12:36

>>5
Good call, I forgot to check that. 2/3 * x^(1/3) + x/3 works fine, however.

Name: Anonymous 2007-12-04 13:22

>>6
Not horizontal in (1,1).

Name: Anonymous 2007-12-04 13:53

It's a circle, radius 1, shifted 1 unit to the right -_-

Name: Anonymous 2007-12-04 15:34

By "horizontal" and "vertical" do you mean dy/dx=0 at 1,1 and dx/dy=0 at 0,0? If so the easiest way would be to set up a some kind of differential equation. dy/dx could easily be some kind of rational expression f(x)/g(x), where f(x)=0 at 1, and g(x)=0 at 0, then just integrate it (the tricky part) to find your function, then ensure that it goes through the points 1,1 and 0,0 (adding c as applicable). Trial and error I think. Just think of different f(x) and g(x) that are 0 and 1 and 0 respectively...

Name: Anonymous 2007-12-04 16:36

>>8
This is correct.
Unless the function needs to be injective.

Name: Anonymous 2007-12-04 17:03

I think this would be a bijective solution:

f(x) =
    -sqrt(-x), if x <= 0
    sqrt(1-(x-1)^2), if 0 < x < 1
    1+(x-1)^2, if x >= 1

Name: Anonymous 2007-12-04 17:42

>>11
Now make it infinitely differentiable on R \ {0}.

Name: Anonymous 2007-12-04 18:45

(x-1)^2 + y^2 = 1

Differentiate

2(x - 1) + 2y (dy/dx) = 0

dy/dx = 2(1 - x) / 2y

dy/dx @ 1,1 = 0/2 = Horizontal

dy/dx @ 0,0 = 2/0 = Vertical

All he said he needed was an equation, not a function. (x - 1)^2 + y^2 = 1 works.

Name: Anonymous 2007-12-04 18:48

(x-1)^2 + y^2 = 1

Differentiate

2(x - 1) + 2y (dy/dx) = 0

dy/dx = 2(1 - x) / 2y

dy/dx @ 1,1 = 0/2 = Horizontal

dy/dx @ 0,0 = 2/0 = Vertical

All he said he needed was an equation, not a function. (x - 1)^2 + y^2 = 1 works.

Name: Anonymous 2007-12-04 18:49

>>14
Hey neat, my internet exploded -_-

Newer Posts
Don't change these.
Name: Email:
Entire Thread Thread List