Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

The implicit equation for a plane perpendicular to a given vector at the origin is $ax + by + cz = 0$.

I can write this in parametric form as $x = t, y = u, z = -\frac{at + bu}{c}$.

The only problem is that I can't use this equation when $c = 0$, and $a$, $b$ and $c$ are each going to be zero at different times so I can't avoid division by zero by shuffling around the coordinates. Is there a way to formulate this plane parametrically without restricting values of $a$, $b$ and $c$?

share|cite|improve this question
None of the three answers given so far is an improvement over what you'd already got (and I don't think there can be any). In each answer, there's a case distinction according to which of $a,b,c$ is non-zero; you can do the same in the solution you already had. (craftsman.don's answer hides the case distinction in "given a normal vector $N$", which also requires a case distinction because no normal vector will work for all cases.) – joriki Oct 4 '11 at 17:58
@joriki: I think the critique in your comment does not apply to my answer (which was written much later). – Did Jan 9 '12 at 16:19
@Didier: You're right, it doesn't :-) That's a nice approach. So I should more specifically say "I don't think there can be any with two parameters". – joriki Jan 9 '12 at 16:27
@joriki: Thanks. It would be nice to make more apparent the connection with (some form of) projective space but, until now, I have been too lazy to do so. – Did Jan 9 '12 at 16:42
up vote 2 down vote accepted

A solution is to parametrize the plane by $\mathbb R^3$. This may seem a strange idea since a plane has dimension $2$ and $\mathbb R^3$ has dimension $3$, but, for the plane of equation $ax+by+cz=0$, such a parametrization is $$ x=\gamma b-\beta c,\quad y=\alpha c-\gamma a,\quad z=\beta a-\alpha b,\quad (\alpha,\beta,\gamma)\in\mathbb R^3. $$ Then, as was to be expected, the parameter $(\alpha,\beta,\gamma)$ corresponding to a given point $(x,y,z)$ in the plane is not uniquely defined. The set of parameters corresponding to $(x,y,z)$ is exactly the affine line $L(x,y,z)$ in $\mathbb R^3$ defined as $$ L(x,y,z)=(\alpha,\beta,\gamma)+\mathbb R\cdot(a,b,c)=\left\{(\alpha+\lambda a,\beta+\lambda b,\gamma+\lambda c)\mid\lambda\in\mathbb R\right\}, $$ where $(\alpha,\beta,\gamma)$ denotes any parameter corresponding to $(x,y,z)$. Note that $L(x,y,z)$ depends on $(x,y,z)$ but not on the choice of $(\alpha,\beta,\gamma)$.

share|cite|improve this answer
Can you please brief how you obtained the parametrization? Thanks. – Kumara Mar 28 '13 at 8:24
Looking for a parametrization with symmetries, since all this is reminiscent of a projective geometry setting. – Did Mar 28 '13 at 8:44
I couldn't get it. Could you please be little elaborate? – Kumara Mar 28 '13 at 9:28
One is looking for vectors in the plane $P$ of equation $ax+by+cz=0$. Obviously $(0,c,-b)$ is in $P$ since $a\cdot0+b\cdot c+c\cdot(-b)=0$. By symmetry, $(-c,0,a)$ and $(b,-a,0)$ are in $P$ as well. And every linear combination of them is in $P$. This is the parametrization in the post. – Did Mar 28 '13 at 10:26

We can find two vectors $u$ and $v$ that are orthogonal to the normal vector of the plane $n = (a, b, c)$. Because not all $a, b, c$ are zero, we can find two components, such at least one of them is not zero. Assume these are $a, b$. The vector $u = (-b, a, 0)$ is orthogonal to $n$ and lies hence in the plane. The vector $v = n \times u = (-ac, -bc, a^2 + b^2)$ is orthogonal to both $n, u$ and lies also in the plane.

Now, the plane can be parametrized as

$$ x = t_1 u + t_2 v$$

where $x$ is an arbitrary vector in the plane given by the scalar parametes $t_1, t_2$.

share|cite|improve this answer

Since $(a,b,c)\ne0$ at least two of the three vectors $$u:=(b,-a,0),\quad v:=(0,c,-b),\quad w:=(-c,0,a)$$ are nonzero, linearly independent, and orthogonal to $(a,b,c,)$. If, e.g., $b\ne0$ then this is the case for $u$ and $v$, and your plane can be parametrized in the form $$x(t_1,t_2):= t_1 u+t_2 v\ .$$

share|cite|improve this answer

let the point set on the plane: P, and the origin: Q

given a normal vector N, then dot(P-Q, N)=0 --------EQU 1

subsitute arbitrary value x,y in to the EQU 1, then you can get a value z. M=(x,y,z) is on that plane.

after that, let U=M - Q, V=N, and W=cross(U,V), then you can get a local frame [U,V,W] of that plane.

any linear combination of U,W can be on the plane. that is, x*U + u*W.

(notice the cross product, cross(U,V), the order is vital. you can change that to cross(V,U) if you want a different handness)

share|cite|improve this answer
I'm afraid I don't understand. Don't I already have $U,V,W$ as $a,b,c$? What is the parametric equation where $(x, y, z)$ is given by $(t, u)$? – jnm2 Oct 4 '11 at 15:28
U,V,W are vectors, not scalar value. You first find out the basis U,V,W, and then parametric equation is Ut + Wu. I make a mistake in the above answer, I am sorry. I have edit it. – craftsman.don Oct 4 '11 at 15:47

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.