advertisement

EECS 203 Spring 2015 Lecture 7 (5/19/15) Page 1 of 4 Review from last time 1. What can you say about the sets A and B if we know that a. A ∪ B = A? b. A − B = A? c. A − B = B − A? 2. Let f : AB and g: BC. a. If f and g are one-to-one, is f ◦ g also one-to-one? b. If f and g are onto, is f ◦ g also onto? c. If g and f ◦ g are onto, is f also onto? 3. Write a formal definition of what it means if function f from AB is onto. 4. Is it true that A B B A? Start on Sequences and Summations (2.4) A sequence is a function from a subset of the set of integers (usually either the set {0, 1, 2, . . .} or the set {1, 2, 3, . . .}) to a set S. We use the notation an to denote the image of the integer n. We call an a term of the sequence. o o Consider the sequence where an=2n. In that case, the sequence, starting with a1 is 2, 4, 6, 8, etc. 1 Consider the sequence where an=𝑛. In that case, what is the sequence? Recurrence Relations A recurrence relation expresses the value of an as a function of previous values. For example, the recurrence relation 𝑎0 = 0, 𝑎1 = 1, 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 for 𝑛 ≥ 2 defines the Fibonacci sequence. That is 0, 1, 1, 2, 3, 5, 8, 13, 21, etc. EECS 203 Spring 2015 Lecture 7 (5/19/15) Page 2 of 4 The sequence is the solution to the recurrence relation. Ideally we’d be able to find it in “closed form” or “closed formula” (that is where we could compute an directly without finding an-1 first). o o Define the even integers (from lowest to highest) as a recurrence relation. Define the factorial function (n!) as a recurrence relation. Recurrence relations hold a special place in computer science. It is often the case that we develop algorithms and then ask ourselves how efficient the algorithm is. And fairly commonly we find ourselves with algorithms where we split the problem into parts and then re-run the algorithm on the parts (recursive programs do this…). We will in fact spend a fair bit of time near the end of this term learning some general techniques for solving recurrence relations (one algorithm turns out to be nearly the same as solving differential equations). Summation of a sequence Sometimes we want to know the sum of a sequence. So the sum of the first N elements (a1+a2+…+aN). We generally write that as 𝑁 ∑ 𝑎𝑖 𝑖=1 Consider the following summation of the sequence ½, ¼, etc. 𝑁 ∑ 𝑖=1 1 2𝑛 Can you find a closed-form solution to that? What is the value as N approaches infinity? Find a closed-form solution for the sum of the sequence 1, 2, 3, etc. EECS 203 Spring 2015 Lecture 7 (5/19/15) Page 3 of 4 Questions 1. What is the value of ∑5𝑖=0(−2)𝑖 ? 2. What is the closed form of the sum ∑𝑛𝑘=0 𝑎𝑘+1 − 𝑎𝑘 ? (This is called a “telescoping series”) 3. Find An example summation Sometimes the telescoping sum trick can be useful where it doesn’t seem like it would be useful. Let’s attack the following summation in two different ways. ∑𝑛𝑘=0 𝑘(𝑘 + 1) Answer one: This is basically 𝑘 2 + 𝑘 summed from 0 to n. By the associative/commutative properties of addition we can write this as ∑𝑛𝑘=0 𝑘(𝑘 + 1) = ∑𝑛𝑘=0 𝑘 2 + ∑𝑛𝑘=0 𝑘. We can start k at 1 rather than 0 (as k and k2 are both 0 when k is 0) and then just use the formulae above for the sum of k and k2. [With a bit of algebra, we can get the sum of k2 from 1 to n to be (1/3)n3 + (1/2)n2 + (1/6)n. So what is ∑𝑛𝑘=0 𝑘(𝑘 + 1)?] EECS 203 Spring 2015 Lecture 7 (5/19/15) Page 4 of 4 Answer two: We can try to manipulate this into a telescoping sum. Note: I’d have never ever come up with this, but it is cool 3𝑘(𝑘 + 1) = [3 + 𝑘 − 𝑘]𝑘(𝑘 + 1) = [(𝑘 + 2) − (𝑘 − 1)]𝑘(𝑘 + 1) = 𝑘(𝑘 + 1)(𝑘 + 2) − (𝑘 − 1)𝑘(𝑘 + 1) = 𝑎𝑘+1 − 𝑎𝑘 when 𝑎𝑘 = (𝑘 − 1)𝑘(𝑘 + 1) And thus we get: And a very relevant series Compound interest is important. If you start with a0 dollars with an interest rate of “r”, you get a1=(1+r)a0. o What is a2 in terms of a1? _________________ in terms of a0? __________________ o What is an in terms of a0?__________________________ What would it look like if in addition, we saved d dollars a year (so a0=d and we add d dollars every year)?