Continued Fractions and Enumeration Sections 7.4 and 8.1-8.3 Continued Fractions Introductory Example: Show 106 = 5+ 19 1+ 1 1 1+ 1 2+ 1 1 1+ 2 = [5;1,1, 2,1, 2] This is known as a (regular or simple) continued fraction. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 2 Continued Fractions Note how this compares to the computations in the Euclidean Algorithm for finding gcd(106,19). Thus we can get the continued fraction representation from the Euclidean Algorithm for rational numbers. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 3 Continued Fractions Note the general pattern, which also works for representing irrational numbers 1 1 x = int1 + frac1 = int1 + = int1 + int 2 + frac 2 1 1 frac = int1 + 1 1 int 2 + 1 2 frac = int1 + 1 1 int 2 + int 3 + frac3 = etc. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 4 Continued Fractions It turns out that all real numbers have continued fraction representations. Further these representations are unique if we do not allow the final &quot;term&quot; to be 1. Also these representations are finite for rational numbers and infinite for irrational numbers. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 5 Continued Fractions Show the square root of 2 equals [1;2,2,2,...] 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 6 Continued Fractions It turns out that square roots of positive integers always have periodic, infinite continued fraction representions. An excellent website for an introduction to continued fractions is in the UT Math Archives at http://archives.math.utk.edu/articles/atuyl/confrac/index. html. Among other information it gives a brief history of continued fractions on which I draw to write the summary below. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 7 Continued Fractions Because of the connection to the Euclidean Algorithm, the time of Euclid is commonly identified as the beginning of the study of continued fractions, but it appears no one used the algorithm this way at that time. From Euclid's time until the 17th century continued fractions arise occasionally in the form of specific examples (for instance they appear in the Indian Aryabhata in the 6th century and the 16th century Italian mathematicians Rafael Bombelli and Pietro Cataldi found the continued fractions for the square roots of 13 and 18 respectively). 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 8 Continued Fractions In the 17th century John Wallis laid the foundation of a general theory of continued fractions, a foundation which Euler, Lambert, and Lagrange greatly advanced in the 18th century. In the 17th century mathematician and astronomer Christiaan Huygens applied continued fractions to the problem of finding rational gear ratios closest to some desired ratio (evidently in order to construct instruments for astronomical study). 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 9 Enumerative Combinatorics (Counting) Broadly speaking combinatorics is the branch of mathematics dealing with order and patterns without regard to the intrinsic properties of the objects under consideration. Among the several topics it includes are graph theory and our current topic, enumeration (counting). 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 10 Enumerative Combinatorics (Counting) There are three kinds of mathematicians. Those who can count, and those who can't. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 11 Enumerative Combinatorics (Counting) It seems universal that elementary introductions to combinatorics emphasize the drawing of tree diagrams as an aid to counting. If they help you, you are welcome to use them; but I have yet to see any value in them. Drawing such a tree becomes awkward when the number of &quot;leaves&quot; becomes larger than perhaps 20. Thus drawing a tree is a practicable approach only for problems so simple as to make such a sophisticated tool superfluous. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 12 Enumerative Combinatorics (Counting) The two main counting rules: The Multiplication Rule states that if one can do a job by doing two tasks one after the other, and there are m ways to do the first task and then n ways to do the second, then there are mn ways to do the whole job. This assumes, of course, that different ways of doing the tasks produce different ways of doing the job and that a suitable choice of ways of doing the tasks can produce every way of doing the job. This result generalizes to more tasks. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 13 Enumerative Combinatorics (Counting) Examples: If your favorite pizza parlor offers three choices of crust, seven meat toppings, and nine vegetable toppings, in how many ways can you order a pizza with one meat topping and one vegetable topping? How many standard TN license plates are there? How many North American telephone numbers are possible? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 14 Enumerative Combinatorics (Counting) The two main counting rules: The Addition Rule states that if one can do a job by doing one or the other (but not both) of two tasks, and there are m ways to do then first task and n ways to do the second, then there are m+n ways to do the whole job. This assumes, of course, that every way of doing the job falls into exactly one of the tasks. This result also generalizes to more tasks. Example: In the pizza example above, how many ways can you order a one-topping pizza? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 15 Enumerative Combinatorics (Counting) A computer system requires a password. A password consists of five uppercase letters. How many passwords are there under the following conditions? No other restrictions No letter repeated No A allowed Exactly one A required At least one A required 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 16 Enumerative Combinatorics (Counting) There are a couple of rules of thumb that often prove useful. When you must choose values for positions under restriction, choose the most restricted position first. Example: How many odd four-digit numbers are there? When objects are supposed to be lined up with certain objects kept together, treat those object that are to be together as a single object (glue them together). Example: A bridal party has six members including the bride and groom. In how many ways can they line up for a photograph if the bride and groom stand next to each other? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 17 Enumerative Combinatorics (Counting) Theorem: An n-set has 2n subsets. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 18 Enumerative Combinatorics (Counting) Definitions A set is an unordered collection of objects without repetition. We denote sets using braces. A sequence of length n on the set S is an n-tuple of elements of S with repetition allowed. We denote sequences using parentheses. For instance (a,b,c,a,a) is a sequence of length five on S={a,b,c,d}. A word (string) of length n on the alphabet S is simply a sequence on length n on S written without braces, spaces, or commas. Thus abcaa is a word of length five on S={a,b,c,d}. A permutation of (finite set) S taken k at a time is a sequence of length k on S without repetition. A permutation of S is a sequence of length |S| on S. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 19 Enumerative Combinatorics (Counting) Equivalent Problems (combinatorial Jeopardy): It is useful in combinatorics to know several counting problems that produce the same answer. Often it is easy to solve an unfamiliar counting problem by recasting in one of these familiar problems. Here are three common answers along with several questions that share them. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 20 Enumerative Combinatorics (Counting) Answer: nk How many sequences of length k on an n-set are there? How many words of length k are there on an alphabet of n letters? In how many ways can you place k labeled balls into n labeled urns? How many functions are there from a k-set to an n-set? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 21 Enumerative Combinatorics (Counting) Answer: P(n,k)=nPk=nk =n(n 1)(n 2)...(n k+1). How many permutations of n objects taken k at a time are there? How many words of length k on an alphabet of n letters are there with no letters repeated? In how many ways can you place k labeled balls into n labeled urns so that no urn gets more than one ball? How many injective functions are there from a k-set to an nset? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 22 Enumerative Combinatorics (Counting) Answer: P(n,n)=n! How many permutations of n objects are there? How many words of length n on an alphabet of n letters are there with no letters repeated? In how many ways can you place n labeled balls into n labeled urns so that every urn gets exactly one ball? How many bijective functions are there from an n-set to an nset. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 23 The Principle of Inclusion and Exclusion The Principle of Inclusion and Exclusion allows us to find the cardinality of a union of sets by knowing the cardinalities of the individual sets and all possible intersections of them The basic version of the Principle of Inclusion and Exclusion is that for two finite sets A and B, it holds that |AB|=|A|+|B| |AB|. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 24 The Principle of Inclusion and Exclusion Example: In a senior class of 100 students 12 failed math, 10 failed chemistry, and 2 failed both. How many failed neither? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 25 The Principle of Inclusion and Exclusion Why does it work? In |AB|=|A|+|B| |AB| we see that |A|+|B| counts everything only in A once and everything only in B once but everything in both A and B twice. By subtracting |AB| we remove the overcount. A Venn Diagram makes this clearer. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 26 The Principle of Inclusion and Exclusion Example: How many numbers from 1 to 100 inclusive are divisible by neither 3 nor 5? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 27 The Principle of Inclusion and Exclusion The result generalizes to three finite sets (in fact it generalizes to any finite number of finite sets): | ABC|=|A|+|B|+|C| |AB| |AC| |BC|+|ABC| Example: How many numbers from 1 to 100 inclusive are divisible by neither 3 nor 5 nor 7? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 28 The Principle of Inclusion and Exclusion Once again a Venn Diagram makes the truth of the Principle of Inclusion and Exclusion for three sets clear. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 29 Permutations and Combinations We already know the basic notation and facts about permutations. The familiar word combinations appears universally in introductory treatments of combinatorics. This is regrettable in that combination is a confusing synonym for subset. For instance a &quot;combination of n objects taken k at a time&quot; means precisely &quot;a subset of size k from a set of size n&quot; or &quot;a k-subset of an nset.&quot;You will improve your own understanding of enumeration and that of your students if you eliminate the word combination from your vocabulary and always use the word subset instead. If you must use a text containing the word combination, teach your students to translate it into subset every time. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 30 Permutations and Combinations We define C(n,k)=nCk= n =the k number of k-subsets of an n-set. Note that this is not the usual definition of C(n,k) by formula. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 31 Permutations and Combinations Example: There are six 2-subsets of {a,b,c,d}, so C(4,2)=6. Note that we have no formula to use to get this result. We actually have to count the subsets. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 32 Permutations and Combinations Theorem k n n! P ( n, k ) n = = = k k! k! k !(n - k )! Combinatorial proof. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 33 Permutations and Combinations Now by this formula C(4,2)=4!/(2!2!)=6. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 34 Permutations and Combinations Examples How many three-topping pizzas are possible if your favorite pizza parlor offers one crust and ten toppings? Watch for possible ambiguity. A club has 10 seniors, 8 juniors, 5 sophomores, and 9 freshmen. In how many ways can the club elect a committee with exactly 2 members from each class? How many words are there with exactly 5 x's and 7 y's? (When are words different?) How many five-card hands are possible from a standard 52card deck? Again watch for possible ambiguity. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 35 Permutations and Combinations Binomial Theorem: n n k n-k n n-k k n ( x + y ) = y = y x x k k k =0 k =0 n The combinatorial proof is much easier than the proof by induction. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 36 Permutations and Combinations The Binomial recurrence (Pascal's Triangle): n n n - 1 - 1 = + k k - 1 k There is an easy and typical but slightly subtle combinatorial proof. There is also a straightforward but unenlightening algebraic proof. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 37 Permutations and Combinations Vandermonde's Formula: m m + n n n = k r k =0 r - k This astonishing result appears impenetrable by algebraic means, but it has a simple combinatorial proof. Skip example 8.47. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 38 Ambiguity Ambiguity is a problem uniquely troublesome in studying enumeration (and thus in studying basic probability). Why is this and what can you do about it? The fundamental enumeration question is, &quot;How many different blivits are there?&quot; This requires you to understand what a blivit is and when two blivits are different. MATHEMATICS CANNOT ANSWER THESE QUESTIONS. THEY BELONG TO THE REAL WORLD. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 39 Ambiguity There are three common reasons ambiguity arises in enumeration problems. The problem assumes real-world knowledge that the student lacks. For example: A lottery requires you to choose five numbers between 1 and 42. How many ways can you make your choice? The problem fails to provide real-world information that the student would have in real life: Examples: How many three topping pizzas can you order at a pizza parlor that offers ten toppings? How many five-card hands can be dealt from a standard 52-card deck? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 40 Ambiguity There are three common reasons ambiguity arises in enumeration problems. Reasonable, well-informed people might disagree about whether a particular object belongs to the set under consideration or whether two objects in the set differ from each other: Example: In how many ways can eight people sit around a round table? 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 41 Ambiguity What do we do about ambiguity? Reduce it by careful wording (but know it will creep in anyway). Make your students aware of it, and encourage them to respond with diligence, courage, and creativity. &quot;I don't believe in no-win situations.&quot; They should seek clarification when possible. Otherwise they should declare a reasonable interpretation and solve the problem accordingly. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 42 Ambiguity What do we do about ambiguity? Allow students the chance to explain their interpretations, knowing that some variant interpretations are legitimate and others are not. When they have a legitimate variant interpretation, recognize and reward it. When they have an illegitimate variant interpretation, explain the real-world reasons that make it illegitimate. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 43 Ambiguity What do we do about ambiguity? Give careful, correct definitions of key mathematical terms. In the rarified atmosphere of mathematics we can eliminate ambiguity, and this is a useful tool in making the real-world concepts. For instance the ambiguous &quot;how many threetopping pizzas are there when ten toppings are available&quot; is completely clarified by asking &quot;how many three-subsets are there of the ten-set of toppings.&quot; We do not want to pose questions this way; but when our students are confused, we need a way to explain to them precisely what we mean. 3/29/2004 <a href="/keyword/discrete-mathematics/" >discrete mathematics</a> for Teachers, UT Math 504, Lecture 10 44