• 沒有找到結果。

Advanced Algebra II Homework 3

N/A
N/A
Protected

Academic year: 2022

Share "Advanced Algebra II Homework 3"

Copied!
4
0
0

加載中.... (立即查看全文)

全文

(1)

Advanced Algebra II Homework 3

Yang Ju-Chen B95201057

2007/03/23 1. Determine Zm ZZn.

Proof.

Let the free Z module on Zm Zn be F; and let the submodule used to create tensor product be K: De…ne the Z module homomorphism f : F ! Z(m;n) by f a; b = ab: Now for any a1; a2; a 2 Zm; b1; b2; b 2 Zn and r 2 Z;

f (a1+ a2; b) a1; b a2; b = f a1+ a2; b f a1; b f a2; b

= (a1+ a2) b a1b a2b = 0;

f a; b1+ b2 a; b1 a; b2 = f a; b1+ b2 f a; b1 f a; b2

= a (b1+ b2) ab1 ab2 = 0;

f a; rb r a; b = f a; rb rf a; b

= a (rb) rab = 0;

f ra; b r a; b = f ra; b rf a; b

= (ra) b rab = 0;

so K ker f: We may identify F= ker f as a submodule of F=K = Zm ZZnby a; b + ker f ! a; b + K = a Zb. It’s clear that Im f = Z(m;n); so F= ker f = Z(m;n); hence

jZm ZZnj jF= ker fj = Z(m;n) = (m; n) :

For any a 2 Zm and b 2 Zn; we have

a Zb = ab1 Z1 2 1 Z1 Z:

Since Zm ZZn is generated from all those elements like above, we conclude that Zm ZZn= 1 Z1

Z: From elementary number theory, there exists x; y 2 Z such that xm + yn = (m; n) : Hence we have

(m; n) 1 Z1 = xm1 Z1 + yn1 Z1 = xm Z1 + y1 Zn = 0 ) 1 Z1 j (m; n) ; in particular,

jZm ZZnj = 1 Z1 (m; n) :

Therefore, jZm ZZnj = 1 Z1 = (m; n) ; Zm ZZn= 1 Z1 Z=Z(m;n):

2. Let M = M1 M2. Prove that M is ‡at if and only if both Mi are ‡at. What can you say if M = i2IMi with general index set I.

Proof.

CautionWhen I is any index set, the symbol i2IMi actually means all the function ' on I such that ' (i) 2 Mi and ' (i) = 0 for all but …nitely many.

Given any exact 0 ! N1 ! N2; say fi : N1 Mi ! N2 Mi; and f : i2I(N1 Mi) !

i2I(N2 Mi) de…ned naturally by fi: Claim ker f = i2Iker fi:

dGiven ' 2 ker f; then f (') = 0 ) f (') (i) = fi(' (i)) = 0; so ' (i) 2 ker fi for all i: Since ' 2 ker f i2I(N1 Mi) ; there is only …nite i such that ' (i) 6= 0; hence ' 2 i2Iker fi:

Now given ' 2 i2Iker fi; say 0 6= ' (i) 2 ker fifor …nite i 2 I0; then fi(' (i)) = f (') (i) = 1

(2)

0 for all i 2 I; hence ' 2 ker f:c

By theorem IV.5.9 on Hungerford, we have

N M = i2I(N Mi)

for any module N: Above claim says that fi is injective for all i 2 I i¤ f is, that is, 0 ! N1 M = i2I(N1 Mi)!f i2I(N2 Mi) = N2 M is exact i¤ 0 ! N1 Mi

fi

! N2 Mi is exact for all i 2 I: Hence M is ‡at i¤ Mi is for all i:

3. Let R be a local ring and M is a …nitely generated ‡at R-module. Then M is free.

In fact, if fx1; :::; xng M such that they form a basis in M=mM over R=m, then it forms a basis of M .

Proof.

Since M is …nitely generated, so is for M=mM; hence dimKM=mM = n < 1 for some n 2 N; where K R=m: Let fx1; :::; xng M such that fx1; :::; xng forms a basis of M=mM;

and let : Rn! M de…ned by

(r1; :::; rn) =X

rixi; ri2 R:

Corollary 1.4.11 says hx1; :::; xni = M; so is surjective. Do the right exact operation m on the exact sequence:

0 ! ker ! Rn! M ! 0 to get the following exact commute diagram:

m ker ! m Rn ! m M ! 0

# # #

0 ! ker ! Rn ! M

Say f : m ker ! ker , g : m Rn ! Rn and h : m M ! M; Snake lemma induces the following exact sequence:

ker h ! coker (f) ! coker (g) ! coker (h) : Now since M is ‡at, exact sequence 0 ! m ! R induces that

0 ! m M ! R M = M is exact, that is, ker h = 0: Therefore,

0 ! coker (f) = ker =m ker ! coker (g) = Rn=mRn= (R=mR)n! coker (h) = M=mM:

But actually (R=mR)n = M=mM because we takefx1; :::; xng as basis of M=mM over R=mR;

so we conclude that ker =m ker = 0; that is, ker = m ker : When R is Noetherian, or M is projective, or more generally, M is …nitely presented (see Lang: Lemma XVI.3.9), ker is

…nitely generated. We can therefore apply Nakayama lemma on the relation ker =m ker = 0 to conclude that ker = 0 and Rn= M .

4. Let 0 ! M1! M2 ! M3 ! 0 be an exact sequence of R-modules. If both M1; M3 are …nitely generated then so is M2.

2

(3)

Proof.

Let f : M1 ! M2 and g : M2 ! M3; then Im f = ker g is …nitely generated, say fxig ; because M1 is. Let fzjg be a generating set of M3; surjection g permits us to choose one yj 2 g 1fzjg for each j; consider X = fxi; yjg : For each x 2 M2; we have for some sj 2 R

g (x) =X

j

sjzj =X

j

sjg (yj) = g 0

@X

j

sjyj

1

A ) x X

j

sjyj 2 ker g;

hence for some ri2 R;

x =X

i

rixi+X

j

sjyj 2 hXi ; so M2 is …nitely generated.

5. * Tensor product commutes with direct limit. That is, lim

!(Mi N ) = (lim

!Mi) N:

Proof.

See Exercise 14 to 20 in Atiyah chapter 2.

6. A ring S is said to be an R-algebra if there is a ring homomorphism R ! S. Show that S is an R-module.

Proof.

Let the ring homomorphism be f; de…ne the operation : R S ! S by r s f (r) s:

For all a; b 2 R, x; y 2 S, we have

a (x + y) = f (a) (x + y) = f (a) x + f (a) y = a x + a y;

(a + b) x = f (a + b) x = f (a) x + f (b) x = a x + b x;

a (b x) = a (f (b) x) = f (a) f (b) x = f (ab) x = (ab) x;

1R x = f (1R) x = 1Sx = x;

so S is an R module.

7. Let S be a ‡at R-algebra and M be a ‡at S-module. Then M is a ‡at R-module.

Proof.

We …rst prove the following useful lemma:

Lemma 1 (Atiyah Exercise 2.15) Let A; B be rings, let M be an A module, P a B module and N an (A; B) bimodule. Then M AN is naturally a B module, N BP an A module, and we have

(M AN ) BP = M A(N BP ) : Proof of Lemma.

For simply, we always use x; y; z; a; b to denote elements in M; N; P; A; B respectively. Any original product will denote without any speci…c symbol. M AN is an A module under A tensor construction, and is module B under new de…ned M AN B :

(x Ay) M AN Bb x A(yb) : Moreover, we have

(a (x Ay)) M AN Bb = ((ax) Ay) M AN Bb = (ax) A(yb)

= a (x A(yb)) = a ((x Ay) M AN Bb) ; 3

(4)

so M AN is an (A; B) bimodule. Similarly, N BP is an (A; B) bimodule under B tensor construction and A N BP :

a A N BP (y Bz) (ay) Bz:

(M AN ) BP is an (A; B) bimodule under B tensor construction and A (M AN ) BP : a A (M AN ) BP (x Ay) Bz (a (x Ay)) Bz:

M A(N BP ) is an (A; B) bimodule under A tensor construction and M A(N BP ): x A(y Bz) M A(N BP )b x A((y Bz) b) :

Fix an z 2 P: The mapping (x; y) ! x A(y Bz) is clearly A linear in x, and x A((ay) Bz) = x A(a A N BP (y Bz)) = a (x A(y Bz)) ;

where the last equality holds by A linearity of A: Hence (x; y) ! x A(y Bz) is A linearity in both x and y; universal property induces a A module homomorphism fz : M AN ! M A(N BP ) such that

fz(x Ay) = x A(y Bz) :

Next, consider the mapping (t; z) ! fz(t) from (M AN ) P into M A(N BP ) ; fz((x Ay) M AN Bb) = fz(x A(yb)) = x A((yb) Bz)

= x A((y Bz) b) = (x A(y Bz)) M A(N BP ) Bb

= fz(x Ay) M A(N BP ) Bb;

fbz(x Ay) = x A(y B(bz)) = x A((y Bz) b) = fz(x Ay) M A(N BP ) Bb;

so (t; z) ! fz(t) is B linear in both t and z (summation homomorphic property is easy to verify). Universal property induces a B module homomorphism f : (M AN ) BP ! M A

(N BP ) such that

f ((x Ay) Bz) = x A(y Bz) : Moreover,

f a A (M AN ) BP ((x Ay) Bz) = f ((a (x Ay)) Bz) = f (((ax) Ay) Bz)

= (ax) A(y Bz) = a (x A(y Bz))

= af ((x Ay) Bz) ;

hence f is an (A; B) bimodule homomorphism. Similarly, we may construct an (A; B) bimodule homomorphism g : M A(N BP ) ! (M AN ) BP such that

g (x A(y Bz)) = (x Ay) Bz:

Clearly f g and g f are identity maps, hence f and g are (A; B) bimodule isomorphism.

Since S is a R algebra, in particular, S is a (R; S) bimodule. Now given any exact R module sequence 0 ! N1 ! N2; that S is R ‡at implies 0 ! N1 RS ! N2 RS is exact. Ni RS is nothing but extension of scalars from R to S; hence S ‡atness of M implies 0 ! (N1 RS) SM ! (N2 RS) SM is exact. By previous lemma and Proposition 1.5.7,

(Ni RS) SM = Ni R(S SM ) = Ni RM;

that is, 0 ! N1 RM ! N2 RM is exact. Therefore, M is a ‡at R-module.

8. * Complete the exercises and incomplete proofs in the note.

4

參考文獻

相關文件

Advanced Algebra II. Homework 1 due

As applications, one have an integrated view of fundamental theorem of finite generated abelian groups and the Jordan canonical form..

Advanced Algebra I. Homework 11 due

It’s clearly an integral extension.. Show that there are only finitely many prime ideals lying

It’s easy to check that m is a maximal ideal, called the valuation ideal. We can show that R is a

The coordinate ring of an affine variety is a domain and a finitely generated k-algebra.. Conversely, a domain which is a finitely generated k-algebra is a coordinate ring of an

The main purpose is to show that the concept of transcendental degree, which is the cardinality of transcendental basis, can be well-defined.. Moreover, transcendental degree is a

The primary decomposition in the al- gebraic side coincide with the decomposition of an algebraic set into union of irreducible components.. We first work on the primary