Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

$s$, $p$ and $c$ are vectors; I need to find $s$ in terms of the other two for:

(1) $| s - c | = 1$

(2) $s = \lambda p$ ( $\lambda$ is a constant )

How can I use the constant $\alpha = (p \cdot c)^2 - p^2 * (c^2 - 1)$?

There may be no solution.

share|improve this question
1  
This is the second problem of the same kind posted within the last hour or so: math.stackexchange.com/questions/33183/… ; are these part of some more general problem, an assignment? What is the context for these problems? –  Arturo Magidin Apr 15 '11 at 19:51
    
The problems are part of some past papers I'm working on for a Maths Methods exam. –  Sorin Cioban Apr 15 '11 at 20:23

2 Answers 2

up vote 1 down vote accepted

The question is not entirely clear. I will try to make a clear question from it, using more or less your notation. It may not be the question you intended to ask!

Let the vectors $p$ and $c$ be given. Find a constant $\lambda$ such that if $s=\lambda p$, then $|s-c|=1$.

Solution: For any vector $v$, $|v|=\sqrt{v\cdot v}$. Since $s=\lambda p$, we want $$(\lambda p -c)\cdot(\lambda p -c)=1$$ Expanding the dot product according to the usual rules, we obtain $$\lambda^2 (p\cdot p) -2\lambda (p\cdot c) +c\cdot c-1=0$$ The above equation is a quadratic equation in $\lambda$ (unless $p$ is the zero-vector).

Solve for $\lambda$ using the Quadratic Formula. Note that there may not be a (real) solution.

share|improve this answer
    
@user6312: "quadratic equation in $\lambda$ (unless $c$ is the zero-vector)" should be "(unless $p$ is the zero vector)". –  Arturo Magidin Apr 15 '11 at 20:33
    
@Arturo Magidin: Thank you! I have made the correction. –  André Nicolas Apr 15 '11 at 20:48

Just as in the other problem you posted, since $s=\lambda\cdot p$, if you know $\lambda$ then you know $s$. If $p=0$, then there is a solution if and only if $|c|=1$, in which case the solution is $s=0$.

Assume $p\neq 0$.

Plugging $s=\lambda p$ into the first equation, you know that $$|\lambda p - c| = 1.$$ Since $|v|^2 = v\cdot v$ for any vector $v$, this gives $$1 = (\lambda p - c)\cdot (\lambda p - c) = \lambda^2 (p\cdot p) - 2\lambda(p\cdot c) + (c\cdot c),$$ or equivalently, that $$(p\cdot p)\lambda^2 - 2(p\cdot c)\lambda + \bigl( (c\cdot c)-1\bigr) = 0.$$ This is a quadratic equation in $\lambda$; if the discriminant $$4(p\cdot c)^2 - 4(p\cdot p)\bigl((c\cdot c)-1\bigr) = 4\alpha$$ (with $\alpha$ the correct version of what you write above; see note at the end) is negative, there are no solutions. If $4\alpha$ is nonnegative, then solving for $\lambda$ gives that $$\lambda = \frac{2(p\cdot c) + \sqrt{4(p\cdot c)^2 - 4(p\cdot p)\bigl((c\cdot c)-1\bigr)}}{2(p\cdot p)}$$ or $$\lambda = \frac{2(p\cdot c) - \sqrt{4(p\cdot c)^2 - 4(p\cdot p)\bigl((c\cdot c) - 1\bigr)}}{2(p\cdot p)}.$$ which yields (up to) two possible solutions for $\lambda$, hence up to two values for $s$.

The constant $\alpha$ in your statement is a rather bad attempt at describing (one fourth of) the discriminant. $p^2$ should be $p\cdot p$ and $c^2$ should be $c\cdot c$. Using $$\alpha = (p\cdot c)^2 - (p\cdot p)\bigl((c\cdot c) - 1\bigr)$$ and simplifying, we can write it as:

  • If $p=0$, then no solutions if $|c|\neq 1$, and $s=0$ is the unique solution if $|c|=1$.
  • If $p\neq 0$ and $\alpha\lt 0$, then no solutions;
  • If $p\neq 0$ and $\alpha\geq 0$, then the solutions are given by $$\lambda = \frac{(p\cdot c) + \sqrt{\alpha}}{p\cdot p},\qquad\text{and}\qquad \lambda = \frac{(p\cdot c) - \sqrt{\alpha}}{p\cdot p}.$$
share|improve this answer
2  
I'm beginning to recognize your posts before I see your name under them -- I saw the question and saw the beginning of your answer and thought, this must be another one of those admirable thorough and patient explanations by Arturo :-) You're really making quite an extraordinary contribution to this site. –  joriki Apr 15 '11 at 23:09
    
@joriki: Thank you kindly! –  Arturo Magidin Apr 15 '11 at 23:47

Your Answer

 
discard

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.