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.

In my perception, using the common sense, is less common, or less probable, to a random line be parallel that not to be, because to be parallel a line needs obey a restrictive rule. But anyone can, using simple probability, obtain that result:

specific line (r) = any line in R2 (e.g. y=0)
amount of random lines (u) = infinite (inf)
amount of parallel lines to r (s) = infinite (inf)
probability of s parallel to r = s / u = inf/inf = undefined

But there is another solution? Maybe using geometric probability, integrals or even empirical results?

[EDIT]

OK. I've understood the zero result and the "not impossible" thing. Thanks for answers and references.

But...

  • Is my initial perception wrong?
  • Is not easier to find a not parallel line instead a parallel?
  • Could someone prove or negate it?
  • If my perception is not wrong someone can calculate how easier is?
  • or maybe I really didnt understand the answers?

[NEW EDIT] So, is it the final answer? - Is my initial perception wrong? - A: No, its correct.

  • Is not easier to find a not parallel line instead a parallel?
  • A: Yes, it is easier.

  • Could someone prove or negate it?

  • A: Yes: '''The probability of finding a parallel line is zero, so the probability of finding a non-parallel line is equal to one. Since 1>0, you have the answer to your question. – Rod Carvalho'''

  • If my perception is not wrong someone can calculate how easier is?

  • A: No, nobody can because it is undefined. (?)

[FINAL EDITION]

Now I realize that the question resumes to: which is the probability of a random real number be equal to a specific other. Thanks for help.

share|improve this question
    
The probability of finding a parallel line is zero, so the probability of finding a non-parallel line is equal to one. Since $1 > 0$, you have the answer to your question. –  Rod Carvalho Sep 14 '12 at 18:57

3 Answers 3

up vote 3 down vote accepted

Suppose that we have a line $\mathcal{L}_1$ defined by

$$\mathcal{L}_1 := \{ (x,y) \in \mathbb{R}^2 \mid y = a x + b \}$$

and another line, $\mathcal{L}_2$, defined by

$$\mathcal{L}_2 := \{ (x,y) \in \mathbb{R}^2 \mid y = \tan (\theta) x + c \}$$

where $\theta$ is an observation of a random variable $\Theta$ uniformly distributed over $[0, \pi]$. Lines $\mathcal{L}_1$ and $\mathcal{L}_2$ will be parallel if $\theta = \tan^{-1} (a)$. However, since $\Theta$ has a continuous distribution, we have that

$$\mathbb{P} \left( \Theta = \tan^{-1} (a) \right) = 0$$

In other words, the probability that a line whose slope is randomly chosen is parallel to a given line is exactly equal to zero.

share|improve this answer
    
Sorry but I cant understand... if the probability is zero then there is no parallel to $\mathcal{L}_1$ ??? –  olivecoder Sep 14 '12 at 18:17
1  
@olivecoder: There is a parallel line, it's just that the probability of getting it is zero. I recommend Terry Tao's explanation of why an event can have zero probability and not be impossible: profiles.google.com/114134834346472219368/buzz/8om1KWEJ5jZ –  Rod Carvalho Sep 14 '12 at 18:23
    
Just wondering why can't we say that the probability is almost zero rather than an absolute zero? In a quick reading of the article, the author is providing an explanation based on the accuracy of the number system and not because the event can't ever happen. –  Emmad Kareem Sep 14 '12 at 21:06
    
@EmmadKareem: OK, suppose that $\Theta$ can take values in a finite subset of $[0, 2\pi]$. Suppose it can take values in $\{ k \frac{2 \pi}{N} \mid k \in \{0,1,\dots,N-1\} \}$. We have a probability mass function where the probability of each angle is $1 / N$. Make $N$ go to infinity, and the probability of each angle goes to zero. If you have an urn with an infinite number of labeled marbles, where each marble is assigned a distinct label in $\mathbb{N}$, the probability of picking a marble labeled "999" is equal to one divided by the cardinality of the naturals, which is zero. –  Rod Carvalho Sep 14 '12 at 21:20
    
This clear now, thanks. –  Emmad Kareem Sep 14 '12 at 21:27

A line in $\Bbb R^2$ is parallel to another line if and only if their slopes are equal, regardless of their intercept.

The probability that the slope $m$ of a random line is equal to the slope $m_f$ of some fixed line is therefore $P(m = m_f) = 0$, since $m_f$ is continuously distributed.

share|improve this answer
    
Sorry but I cant understand... if the probability is zero then there is no parallel to f line ??? –  olivecoder Sep 14 '12 at 18:17
    
A probability of zero is not the same thing as an impossible event. For example, take a uniform random variable $u$ continuously distributed in $(0,1)$. What is the probability that $u = .56721$? The answer is $P(u=.56721) = 0$. In fact, the probability of $u$ assuming any value in the interval is zero. However, any sampling of $u$ must come up with a value in the interval, so clearly probability zero does not imply impossibility. –  Arkamis Sep 14 '12 at 18:23

The answer depends on the random distribution accordingto which the lines are produced. But if the probability of a random line being parallel to a given line $L$ is assumed not to depend on $L$ (a resonable symmetry assumption), then the assumption of a positive probability $p>0$ leads to a contradiction: If $p>0$ then there exists a natural number $n$ such that $n>\frac1p$. It is easy to construct $n$ differnt lines $L_1,\ldots, L_n$ through a single point $A$. The probability of a random line $L$ to be parallel to $L_i$ is assumed to be $p$ for all $i$. Since the events $L||L_i$ and $L||L_j$ are mutually exclusive, we obtain that the probability that $L$ is parallel to any of the $L_i$ is $p+p+\ldots+p$ with $n$ summands. This produces a total probability of $np>1$ - contradiction. We conclude that the assumption $p>0$ is wrong, hence $p=0$.

share|improve this answer

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.