PUMaC 2023 · 加试 · 第 1 题
PUMaC 2023 — Power Round — Problem 1
题目详情
Problem 1.1.4. Show that the set of odd integers, as a subset of Z , is not an ideal. (Hint: which property does this set not satisfy?) Sometimes, we want to specify an ideal of R without having to explicitly list all of the elements. In particular, we only need to specify a subset of the elements of the ideal, knowing that our ideal satisfies the properties in the definition. For instance, note that any ideal that contains 2 also must contain the even integers. Indeed, note that if 2 lies in an ideal I, then so does every even integer, since each even integer is equal to 2 times some other integer. As the even integers are an ideal, it thus makes sense to describe the ideal of even integers as the smallest ideal that contains 2 : that is, every other ideal containing 2 contains the even integers, and the even integers are precisely the set of integers which are a multiple of 2 . These notions, of each even integer being a multiple of 2 , and of the even integers being the smallest such ideal containing 2 , motivates the notion of generators of an ideal. Definition 1.1.4. We say that an ideal I is generated by a subset of elements S ⊂ I if n P every element i ∈ I can be written in the form i = r s , for some positive integer n, and j j j =1 elements s , s , . . . , s ∈ S and r , r , . . . , r ∈ R. 1 2 n 1 2 n Similarly, given a ring R and elements s , s , . . . , s , we let ⟨ s , s , . . . , s ⟩ be the set of 1 2 n 1 2 n n P elements of the form i = r s for elements r , r , . . . , r ∈ R. We can also substitute a j j 1 2 n j =1 n P set, letting ⟨ S ⟩ be the set of elements in R of the form i = r s for some positive integer j j j =1 n, and elements s , s , . . . , s ∈ S, r , r , . . . , r ∈ R. 1 2 n 1 2 n With the notation above, the set of even integers, 2 Z , can also be written as ⟨ 2 ⟩ . Note that the set of even integers is an ideal. This happens more generally, as follows. Proposition 1.1.5. Given a subset S ⊂ R, the set ⟨ S ⟩ is an ideal. ′ Proof. For the first condition, suppose that we have two elements in ⟨ S ⟩ , say i and i . ′ ′ ′ Then, there are positive integers n, m and elements s , s , . . . , s , s , s , . . . , s ∈ S and 1 2 n 1 2 m n m P P ′ ′ ′ ′ ′ ′ r , r , . . . , r , r , r , . . . , r ∈ S such that i = r s and i = r s . Then, their sum 1 2 n j j m 1 2 j j j =1 j =1 n m P P ′ ′ ′ is equal to i + i = r s + r s , which is of the given form. Notice that we can j j j j j =1 j =1 ′ further simplify this expression if we know that some of the s and s are equal, using the j j distributive property. n P For the second condition, given an element i in ⟨ S ⟩ , we can write it as r s for some j j j =1 positive integer n, s , s , . . . , s ∈ S and r , r , . . . , r ∈ R. But then, for each element 1 2 n 1 2 n n P r ∈ R, observe that ri = rr s , which is also in ⟨ S ⟩ . This finishes the proof of the j j j =1 proposition. Definition 1.1.6. We call the set ⟨ S ⟩ the ideal generated by S. We now wish to generalize the notion of a prime from Z . Rather than thinking about elements as being primes, we want to think about ideals. The main behavior we want to capture is the fact that, given a prime number p, if ab lies in p Z , then either a or b lies in it. Contrast this with 6 Z , for instance: 2 · 3 lies in 6 Z , but 2 , 3 do not lie in 6 Z . Definition 1.1.7. An ideal I of a ring R is said to be prime if it is a proper ideal, and furthermore, for all a, b ∈ R, ab ∈ I implies that either a ∈ I or b ∈ I. For instance, the ideal ⟨ 2 ⟩ ⊂ Z is a prime ideal, since ab ∈ ⟨ 2 ⟩ if and only if ab is an even integer; but notice that one of a, b must be even as well.
解析
Problem 1.2.5. Show that the set of elements Z [ − 13] , of the form a + b − 13 , for a, b ∈ Z , while an integral domain, is not a UFD, and therefore not a PID. Solution. We first verify that this is a ring; with the operations inherited from C , we just need to show that it contains the multiplicative and additive identities (which is clear, as 0 , 1 ∈ Z ⊂ √ Z [ − 13]), and is closed under addition, multiplication, and additive inverse. √ √ √ √ But given a + b − 13 , c + d − 5 ∈ Z [ − 13] their sum is ( a + c ) + ( b + d ) − 13 , their product √ √ √ is ( ab − 13 bd ) + ( ad + bc ) − 13 , and the additive inverse of a + b − 5 is ( − a ) + ( − b ) − 13 , which √ √ for a, b, c, d ∈ Z lie in Z [ − 13] since Z is a ring. Hence, Z [ − 13] is a ring. √ √ Furthermore, it is an integral domain. Indeed, if ( a + b − 13)( c + d − 13) = 0 , then we have √ √ that ( ac − 13 bd ) + ( ad + bc ) − 13 = 0 , or that 13 bd − ac = ( ad + bc ) − 13 . But the right-hand side √ is not an integer unless ad + bc = 0 , and so ac − 13 bd = 0 too. Now, suppose that a + b − 13 ̸ = 0 , − bc so we have one of a, b is nonzero. If it is a, then notice that ad + bc = 0 implies that d = , and a 2 13 b c 2 2 2 2 2 2 so ac − 13 bd = ac + = 0 , or that a c + 13 b c = 0 , or ( a + 13 b ) c = 0 . If a + 5 b = 0 , then as a squares of integers are positive, we would need a, b = 0 , contradiction. Thus, c = 0 , and therefore 2 − ad − a d d = 0 . Similarly, if b ̸ = 0 , we have that c = , and so ac − 13 bd = − 13 bd = 0 , or that b b 2 2 a d + 13 b d = 0 , where the same argument lets us conclude that c = d = 0 . Hence, R is an integral domain. √ √ To show it is not a UFD, consider the factorizations 14 = 2 · 7 = (1 + − 13)(1 − − 13) . On the √ √ one hand, notice all the elements involved are irreducible. Indeed, if 2 = ( a + b − 13)( c + d − 13) , then ac − 13 bd = 2 and ad + bc = 0 . But then either a or b is nonzero. But notice then that √ √ 2 2 2 ( a − b − 13)( c − d − 13) = 2 as well, meaning that multiplying these implies that ( a + 13 b )( c + 2 2 2 2 2 2 2 13 d ) = 4 . But notice that a + 13 b cannot equal 2 . Hence, either a + 13 b = 1 or c + 13 d = 1 , √ √ meaning that one of a + b − 13 , c + d − 13 is ± 1 , and so 2 is irreducible. By the same logic, we have that 7 is irreducible. √ √ √ √ As for 1 ± − 13 , notice that if (1 + − 13) = ( a + b − 13)( c + d − 13) , then ac − 13 bd = √ √ √ 1 , ad + bc = 1 , and so similarly we find that (1 − − 13) = ( a − b − 13)( c − d − 13) , so multiplying 2 2 2 2 2 2 2 2 these together yields ( a + 13 b )( c + 13 d ) = 14 . But a + 5 b , c + 5 d are positive, and cannot √ √ √ equal 2 or 7 , so one of these is 1 , and so one of a + b − 13 , c + d − 13 is ± 1 , so 1 ± − 13 is irreducible. √ √ However, notice that 1 ± − 13 are not divisible by 2 , since for any a + b − 13 , we have √ √ √ √ 2( a + b − 13) = 2 a +2 b − 13 , where a, b are integers, and so we would need 1 ± − 13 = 2 a +2 b − 13 , √ √ or (2 a − 1) = ( ± 1 − 2 b ) − 13 . But this cannot happen as − 13 is not rational. In particular, we have two factorizations where one of them is not simply a rearrangement of √ the other, or has different unit multiples. Thus, Z [ − 13] is not a UFD, ergo not a PID. 1.3 Product Rings, Quotient Rings and More Examples ′