You are on page 1of 4

INTRO TO PROOF JOURNAL THREE

MARCUS BEADLE Abstract. A metric set has the condition that it is symmetric, non-negative, zero if there is no transformation, and follows the triangle inequality. The following three theorems suggest that certain functions such as the discrete metric are metric on their respective sets and will be proven by showing the functions of these sets meet these conditions or properties therefore are indeed metric.

1. Introduction In 1676 the term geometria situs, which means geometry of position, was introduced by Leibniz. His idea was that distance rather than magnitude was the most important factor of the study and application of geometry. This foundation was built upon by Leonhard Euler sixty years later when he introduced his famous Koenigsberg Bridge Problem and gave the world the rst real-world application to the study of topology. This problem is described by in the Principles of Topography[1]. Some mathematicians[2] consider the most important example of topological spaces to be metric space. They have more structure than many of the other, more abstract topological spaces and also the typical spaces for mathematical analysis. In this journal, I will explore some of the more popular metric functions and prove that they are in fact metric. 2. Definitions Many of the ideas presented in this journal are commonplace in normal life and are used without cognizance. For example, if one walks one block down mulberry street it is obvious that the distance to get back to the start is equal to the distance traveled which is a manifestation of the symmetry property. This journal will be proving that certain functions are metric on specic sets which is slightly more abstract than simply distances traveled while walking. But comparing the distance between elements or points in a set can be easily compared to a physical distance and will be used as an example throughout the journal. We will begin with the rst denition. Denition 2.1. A Metric To be considered a metric on a set X for a function d : X X R such that all a, b, c X, the four following conditions must be met: The rst condition will be called the non-negative property. 1) d(a, b) 0. It means that the function is a non-negative number, much like the magnitude of a vector or the absolute value. This property shows that only distance matters, not direction. The next property will be called the zero property. 2) d(a, b) = 0 if and only if a = b. This means that the magnitude of the function will can only be zero if both terms in the function are zero. For example, assume a rock is thrown
Date: 12 April 2012.
1

MARCUS BEADLE

from point A and it lands at point B. If the rock traveled a distance of zero, then the rock has not moved meaning point A is point B. Otherwise the distance would be greater than zero. The third condition is called the symmetry property. 3)d(a, b) = d(b, a). Consider two points, A and B, on a line. The distance from point A to point B is equal to the distance from point B to point A. The symmetry property can only be true if both the non-negative property and the zero property are also true. The nal condition is called the triangle property. 4)d(a, b) + d(b, c) d(a, c). This is a property of geometry suggesting that if you have three points on a plane-A, B, and C-then the distance between points A and B plus the distance between points B and C will always be greater than or equal to the distance from point A to point C.

3. Theorems, Proofs, and Examples Consider the Euclidean Distance formula d(x, y) = (x1 y1 )2 + (x2 y2 )2 + ... + (xn yn )2 where xn and yn are points on a line. When there are only two points we can write d(x, y) = (x y)2 = |x y|. The absolute value is added because (x y)2 cannot be negative since the dierence of any two numbers squared is non-negative. Here we are looking for magnitude or a distance between the points a and b. This distance function is obviously a metric. Some mathematicians[2] are so certain that they leave such an easy proof to the reader. For those such as the students of the Hasse Prize Award winning, combinatorial mathematical ladder game co-creating Dr. Steven Dougherty[3], who take pride in their proofs, each step has been provided. Theorem 3.1. The set X = N of d(x, y) = |x y| is a metric. Proof. It will be proven that this set is a metric by showing it satises the conditions mentioned in the denition. The rst condition that will be shown is the condition of nonnegativity. Since d(x, y) = |x y| then it satises d(a, b) 0. No matter what values of x or y are used, because the set is an absolute value, it will always be greater than or equal to zero. Now I will show that the function satises the second condition, the zero property. Assume d(x, y) = 0. Then |x y| = 0. This is only true for x = y, therefore the set satises the zero property. The third condition, the symmetric property, will now be shown to be true for this set. Assume x y = q for some q N. Then d(x, y) = |x y| = |q| = q. In turn, d(y, x) = |y x| = | q| = q. Therefore the condition d(a, b) = d(b, a) is true. Finally I will show that the triangle inequality is true. Let d(x, z) = |x z|. Then |x z| = |x y + y z|. Because all terms are absolute values we can say |x z| |x y| + |y z|. Therefore d(a, c) d(a, b) + d(b, c). Because all four conditions of a metric have been shown to be true, it has been proven that X = N of d(x, y) = |x y| is a metric. Because the formula here only accounts for two points on a line this is a very simple case. However, the same reasoning from the proof above can be used to show that any form of the
n

Euclidean distance formula, d(x, y) =


i=1

|xi yi |2 , is a metric.

Lets dene d(x, y) of the arbitrary set X to be 0 when x = y and 1 when x = y. This is called the discrete metric. Example 3.1.3[1] suggests that it may be uninteresting but is still

INTRO TO PROOF JOURNAL THREE

important because it shows that any set can be assigned a metric. However, the author of that example fails to show how the subsets of the discrete metric are both open and closed which in itself is very interesting. This idea is explained in example 4.3.6[4]. Since open versus closed metric sets will not be discussed in this journal it will be left to the reader to see why this makes sense. It will only be shown that the discrete metric is actually a metric. Theorem 3.2. The set X = R of d(x,y) = 0 for x = y is a metric. 1 for x = y

Proof. Just as it was done for the previous theorem these proving this theorem true will be accomplished by showing each of the four conditions of a metric are true for this set beginning with the non-negative property. Since the discrete metric only has two possible values, 1 and 0, d(x, y) is always greater than or equal to zero. This satises the non-negative condition. Now it will be shown that the zero condition applies. Assume x = y. Then d(x, y) = d(y, x) which satises the condition of symmetry for that case. But when x = y then d(x, y) = 1 and d(y, x) = 1 so the set is symmetric when x = y. Since in both possible cases d(x, y) = d(y, x) the set satises the condition of symmetry. Since the rst three conditions are more than obvious, many mathematicians[2] would just note the rst three conditions as being trivial or straight forward and continue to the triangle inequality. This will be shown by assuming the condition is not true which leads to a contradiction. Assume d(x, z) > d(x, y) + d(y, z). Since 1 and 0 are the only possible vales then d(x, z) must equal 1 and d(x, y) + d(y, z) must equal 0. For d(x, z) = 1 then x = z. For d(x, y) + d(y, z) = 0 then x = y and y = z which implies x = z. It was assumed x = z, therefore the triangle inequality is true for this set. By showing that all four conditions are true for the set, it has been proven that the discrete metric is actually a metric. For the reader who has had their rst encounter with the discrete metric, it is easy to see why one my suggest it is of little interest. If this were applied to Euclidean distances it would tell you either there is a distance when 1 or there is not a distance. when 0. However in more advanced applications this simple piecewise function can evoke some very complex ideas. Now we shall look at a set X described by the function d((x, y), (z, w)) = (x z)2 + (y w)2 . Here we will see that a set X = R R can be described as a distance in topological space as a metric. It is important not to be fooled by the extra variables in this function. It works just as the Euclidean distance function did in the rst theorem and will be proven the same way as the last two. Theorem 3.3. The Set X = R R of d((x, y), (z, w)) = (x z)2 + (y w)2 is a metric.

Proof. I will begin by showing (x z)2 + (y w)2 satises the non-negative property by working from the inside of the function outward. The statements (x z)2 and (y w)2 are non-negative since the square of number is always non negative. Then their sum is non-negative because the sum of any two non-negative numbers is non-negative. Then the function as a whole must always be non-negative since the square root of any non-negative number is non-negative. Therefore this set satises the non-negative property. To show the zero property is true d((x, y), (z, w)) = 0. This is true if and only if (x z) = 0 and (y w) = 0. This implies x = z and y = w so (x, y) = (z, y). Therefore the set satises the zero property. To show the condition of symmetry, assume x z = r and

MARCUS BEADLE

y w = v for some r, v R. Then d((x,y),(z,w)) = (x z)2 + (y w)2 = r2 + v 2 and d((x, y), (z, w)) = (z x)2 + (w y)2 = (r)2 + (v)2 = r2 + v 2 because the square of a non-positive number is non-negative. Since d(a, b) = d(b, a), this set satises the condition of symmetry. Now it will be shown that this set satises the triangle inequality property. Since d((x, y), (s, t))+d((s, t), (z, w)) = (x s)2 + (y t)2 + (s z)2 + (t w)2 for some s, t R and (x s)2 + (y t)2 + (s z)2 + (t w)2 (x z)2 + (y w)2 = d((x, y), (z, w)) the set satises the triangle inequality. Since all four conditions of a metric have been met for this set, then it has been proven that X = R R of d((x, y), (z, w)) = (x z)2 + (y w)2 is a metric.

References
[1] Croom, Fred H. Principles of Topology Saunders College Publishing: Philadelphia (1989) [2] Krants, Steven G. A Guide to Topology Mathematical Association of America: Washington (2009) [3] Dougherty, Steven T. https://sites.google.com/site/professorstevendougherty/home. Accessed April 2012 [4] Searcoid, Micheal O. Metric Spaces Springer: London (2007)

You might also like