10월, 2018의 게시물 표시

Basic properties of ring homomorphisms

R,\ S : ring. \phi : R \rightarrow S be ring homomorphism. Let A be subring of R and B be ideal of S. 1. For any r \in R and any positive integer n, \phi(nr) = n\phi(r) and \phi(r^{n}) = (\phi(r))^{n} By operation-preserving property of ring homomorphism, following equations must be satisfied. \phi(r + \dots + r) = \phi(r) + \dots + \phi(r) \phi(r \cdot r \dots r) = \phi(r) \cdot \phi(r) \dots \phi(r) 2. \phi(A) = \{\phi(a) \mid a \in A\} is subring of S. Since that A is subring of R, a - b \in A and ab \in A,\ \forall a,\ b \in A. Hence, \phi(a - b) = \phi(a) - \phi(b) \in \phi(A) and \phi(ab) = \phi(a)\phi(b) \in \phi(A),\ \forall a,\ b \in A). \therefore \text{By subring test, } \phi(A) \text{ is subring of }S. 3. If A is an ideal and \phi is onto S, then \phi(A) is an ideal. \phi is onto S implies that \phi(r) = s ,\ \exists r \in R,\ \forall s \in S or equivalently, \phi(A) = S. A is ideal of R impl...

Reducibility over Q implies reducibility over Z

Q : ring of rational number (which is field) Z : ring of integer (which is integral domain, not a field) content of polynomial : greatest common divisor of the coefficients of polynomial (must be nonzero polynomial) Suppose f(x) \in Z[x] and f(x) = g(x)h(x), where g(x),\ h(x) \in Q[x] and 1 \leq deg\ g(x),\ h(x) < deg\ f(x). Also, we can assume that f(x) is primitive polynomial. Since that dividing both sides with the content of f(x), say c_{f(x)}, derives  \frac{f(x)}{c_{f(x)}} = \frac{g(x)h(x)}{c_f(x)} Which makes polynomial in left side primitive, but does not affect degree of g(x) nor h(x). Hence, f(x) can be primitive in any situation. Let a be the LCM(Least Common Divisor) of the denominators of the coefficients of g(x), and b be the LCM of the denominators of the coefficients of h(x). Then, abf(x) = ag(x) \cdot bh(x)\ (ag(x),\ bh(x) \in Z[x]). Let c_{1} be the content of ag(x), and c_{2} be the content of bh(x). ...

Reducibility test for degrees 2 and 3

F : field Reducibility test for degrees 2 and 3 is as follows, f(x) \in F[x] \text{ and } deg\ f(x) = 2\text{ or }3 \implies f(x)\text{ is reducible over }F\text{ iff }f(x)\text{ has a zero in }F Suppose that f(x) = g(x)h(x), where g(x),\ h(x) \in F[x]. Since that, polynomials of zero degree are the only unit over integral domain , forms containing zero polynomial are not our concern. Hence, 1 \leq g(x),\  h(x) < deg\ f(x) Since, in integral domain , deg f(x) = deg\ g(x) + deg\ h(x) and deg\ f(x) is 2 or 3, at least one of g(x) and h(x) has degree 1. Say g(x) = ax + b. Then, clearly, -a^{-1}b is a zero of g(x) and therefore a zero of f(x). Conversely, suppose that f(a) = 0, a \in F. Then by the Factor Theorem, we know that x - a is a factor of f(x) and, therefore f(x) is reducible over F.

On the form of ring homomorphism from Z_{n} to itself.

\phi : Z_{n} \rightarrow Z_{n}, x \mapsto ax, where a^{2} = a is ring homomorphism. Let \phi(1) = a. Then a = \phi(1) = \phi(1)\phi(1) = a^{2}. Let x \in Z_{n}. Then \phi(x) = \phi(1 + \dots + 1) = \phi(1)x = ax. Hence, our original conjecture is true.

Kernels are ideals

R,\ S : ring. \phi : R \rightarrow S be ring homomorphism. Let Ker(\phi) = \{r \in R \mid \phi(r) = 0 \}. Since that \phi(a - b) = \phi(a) - \phi(b) = 0_{R} + 0_{R} = 0_{R}. Hence, a - b \in Ker(\phi), \forall a,\ b \in Ker(\phi) Also, by as follows, \phi(rx) = \phi(r) \phi(x) = \phi(r) \cdot 0_{R} = 0_{R},\ \ \forall r \in R,\ \forall x \in Ker(\phi) \phi(xr) = \phi(x) \phi(r) = 0_{R} \cdot \phi(r) = 0_{R},\ \ \forall r \in R,\ \forall x \in Ker(\phi) rx,\ xr \in Ker(\phi),\ \ \forall r \in R,\ \forall x \in Ker(\phi) is trivial. \therefore \text{By ideal test, } Ker(\phi)\text{ is ideal of }R

Every zero-divisors in Z_{p^{n}} is a nilpotent element, where p is a prime.

\mathbb Z_{p^{n}} : set of integers modulo p^{n} nilpotent : a^{n} = 0,\ \exists n \in \mathbb Z^{+} Note that a \in \mathbb Z_{n} is zero-divisor if and only if gcd(a, n) = d,\ (d > 1). Since that d \mid p^{n}, it implies d = p^{k},\ (k < n). a = ld = lp^{k} a^{n} = (lp^{k})^{n} = (l)^{n}(p^{k})^{n} = (l)^{n}(p^{n})^{k} = 0.

Order of finite field is p^{n}, where p is prime.

F : finite field. char(F) : characteristic of F We already know that (F,\ +) is the group under the addition. Since that every field is integral domain,  characteristic of F must be 0 or some prime p. Characteristic of finite field could not be 0. Since that, by \textit{Lagrange's Theorem}, the order of each element of the group divides the order of the group if the order of the group is finite. So, let's consider the case when char(F) is some prime p. Then, 0 = 1 \cdot p which implies |1| = p on (F,\ +) and p \mid |F|. Let's assume that there is some other prime q, where q \mid |F|. By the \textit{Cauchy's Theorem}, if q \mid |F|, then |x| = q, \exists x \in F. If char(F) = p, then x \cdot p = 0 and x \cdot q = 0 must be satisfied. But, from the fact that gcd(p,\ q) = 1, by Bezout's identity, there exists some integers a, b which satisfies ap + bq = 1. By multiplying x on both sides of equation yiel...

If G is self-complementary graph, then 4 \mid |V_{G}| or 4 \mid |V_{G}| - 1

G : undirected simple finite graph, \bar{G} : complementary graph of G. self-complementary : G \cong \bar{G} Between complementary graph, |E_{G}| + |E_{ \bar{G}}| = \frac{|V_{G}|(|V_{G}| - 1)}{2}should be satisfied. If G \cong \bar{G}, then |E_{G}| = |E_{\bar{G}}|. Hence, with substitution, 2|E_{G}| = \frac{|V_{G}|(|V_{G}| - 1)}{2} \\[10pt] 4|E_{G}| = |V_{G}|(|V_{G}| - 1) \\[10pt] If |V_{G}| is even integer, then |V_{G}| - 1 is odd integer. Converse is also trivially true. \therefore 4 \mid |V_{G}| \ \text{or} \ 4 \mid |V_{G}| - 1

The only self-complementary cycle graph is C_{5}

G : undirected simple finite graph, \bar{G} : complementary graph of G. C_{n} (n \geq 3) : n-cycle graph, which is the graph itself is a cycle. To be G self-complementary, G \cong \bar{G} should be satisfied. Hence, G and \bar{G} must have same number of edges. Which implies that |E_{G}| = |E_{ \bar{G}}| = \frac{1} {2} \cdot \frac{|V_{G}|(|V_{G}| - 1)}{2} (Note that |V_{G}| = |V_{\bar{G}}| by definition of complementary graph). From the fact that C_{n} is cycle graph, |E_{C_{n}}| = |V_{C_{n}}| = n and |E_{C_{n}}| = |E_{\bar{C_{n}}}|, |E_{C_{n}}|  + |E_{\bar{C_{n}}}| = \frac{|E_{C_{n}}|(|E_{C_{n}}| - 1)}{2} \\[10pt] 2n = n(n-1) \\[10pt] n(n - 5) = 0 \\[10pt] \therefore n = 0, 5 Since that n \geq 3, the only answer is 5.

The intersection of any set of ideals of a ring is an ideal.

R : ring, D = \{I_1, \dots , I_n \} : arbitrary set of ideals of R. Let's show that S \subseteq D, \bigcap \limits_{I_{i} \in S}{I_{i}} is ideal of R. Since that D satisfies a - b \in I_{i} (\forall a, \forall b \in \bigcap \limits_{I_{i} \in S}{I_{i}})(\forall I_{i} \in S) and ar, ra \in I_{i} \forall I_{i} \in S. Hence, by ideal test, it is trivial.

Proper ideal I of commutative ring with unity R, is unique maximal ideal of R if u is unit of R\ \forall u \in I^{c} \cap R.

R : commutative ring with unity, I : proper ideal of R. Assume that a \in R is unit \forall a \notin I. Let's first show that I should be maximal ideal of R. We already know that ideal with unit could never be proper ideal.  So we can conclude that I never contains any unit element in R. Since that every element outside of I is unit, any ideal of R which properly containing I, must have unit of R. Which implies I is maximal ideal by definition. Next, we need to prove that I is the unique maximal ideal of R. Obviously, any proper ideal of R must be in I. Therefore, I is the only maximal ideal.

Quotient ring of principal ideal domain is principal ideal.

R : principal ideal domain, I : ideal of R. Since that R is the principal ideal domain, every ideal of R is principal ideal. Let such ideal be H. Also, let J =\{a \in R \mid a + I \in H\} (Could be interpreted as the set of the representatives of cosets in R/I). Let's prove that J is the ideal of R. Since that ring must have the additive identity, (which implies that the ring is non-empty set) we can say that 0 + I \in H, 0 \in J. Also, a, b \in J, \forall a + I, b + I \in H which implies (a + I) - (b + I) = (a - b) + I \in H. So, a - b \in J. Since that H is the ideal of R/I, ar + I, ra + I \in H, \forall a \in J, \forall r \in R which implies ar, ra \in J. Therefore, we can conclude that J is the ideal of R. Next, let's prove H = <j + I>, \exists j \in J. (\Longleftarrow)     Obiviously, <j + I> \subseteq H, \exists j \in J. (\Longrightarrow)     If a + I \in H, then a \in J. ...