Difference between revisions of "Bezout's Lemma"
Hashtagmath (talk | contribs) (→See also) |
Mathpro1441 (talk | contribs) (→Proof) |
||
Line 10: | Line 10: | ||
Since <math>\gcd(x_1,y_1)=1</math>, <math>\text{lcm}(x_1,y_1)=x_1y_1</math>. So <math>\alpha=y_1</math> is smallest positive <math>\alpha</math> for which <math>x\alpha\equiv 0\pmod{y}</math>. Now if for all integers <math>0\le a,b<y_1</math>, we have that <math>x_1a\not\equiv x_1b\pmod{y_1}</math>, then one of those <math>y_1</math> integers must be 1 from the [[Pigeonhole Principle]]. Assume for contradiction that <math>x_1a\equiv x_1b\pmod{y_1}</math>, and WLOG let <math>b>a</math>. Then, <math>x_1(b-a)\equiv 0\pmod {y_1}</math>, and so as we saw above this means <math>b-a\ge y_1</math> but this is impossible since <math>0\le a,b<y_1</math>. Thus there exists an <math>\alpha</math> such that <math>x_1\alpha\equiv 1\pmod{y_1}</math>. | Since <math>\gcd(x_1,y_1)=1</math>, <math>\text{lcm}(x_1,y_1)=x_1y_1</math>. So <math>\alpha=y_1</math> is smallest positive <math>\alpha</math> for which <math>x\alpha\equiv 0\pmod{y}</math>. Now if for all integers <math>0\le a,b<y_1</math>, we have that <math>x_1a\not\equiv x_1b\pmod{y_1}</math>, then one of those <math>y_1</math> integers must be 1 from the [[Pigeonhole Principle]]. Assume for contradiction that <math>x_1a\equiv x_1b\pmod{y_1}</math>, and WLOG let <math>b>a</math>. Then, <math>x_1(b-a)\equiv 0\pmod {y_1}</math>, and so as we saw above this means <math>b-a\ge y_1</math> but this is impossible since <math>0\le a,b<y_1</math>. Thus there exists an <math>\alpha</math> such that <math>x_1\alpha\equiv 1\pmod{y_1}</math>. | ||
− | Therefore <math>y_1|(x_1\alpha-1)</math>, and so there exists an integer <math>\beta</math> such that <math>x_1\alpha - 1 = y_1\beta</math>, and so <math>x_1\alpha + y_1\beta = 1</math>. Now multiplying through by <math>g</math> gives, <math>gx_1\alpha + gy_1\ | + | Therefore <math>y_1|(x_1\alpha-1)</math>, and so there exists an integer <math>\beta</math> such that <math>x_1\alpha - 1 = y_1\beta</math>, and so <math>x_1\alpha + y_1\beta = 1</math>. Now multiplying through by <math>g</math> gives, <math>gx_1\alpha + gy_1\beta = g</math>, or <math>x\alpha+y\beta = g</math>. |
Thus there does exist integers <math>\alpha</math> and <math>\beta</math> such that <math>x\alpha+y\beta=g</math>. | Thus there does exist integers <math>\alpha</math> and <math>\beta</math> such that <math>x\alpha+y\beta=g</math>. |
Revision as of 11:06, 26 June 2019
Bezout's Lemma states that if and are nonzero integers and , then there exist integers and such that . In other words, there exists a linear combination of and equal to .
Furthermore, is the smallest positive integer that can be expressed in this form, i.e. .
In particular, if and are relatively prime then there are integers and for which .
Proof
Let , , and notice that .
Since , . So is smallest positive for which . Now if for all integers , we have that , then one of those integers must be 1 from the Pigeonhole Principle. Assume for contradiction that , and WLOG let . Then, , and so as we saw above this means but this is impossible since . Thus there exists an such that .
Therefore , and so there exists an integer such that , and so . Now multiplying through by gives, , or .
Thus there does exist integers and such that .
Now to prove is minimum, consider any positive integer . As we get , and as and are both positive integers this gives . So is indeed the minimum.
Generalization to Principal Ideal Domains
Bezout's Lemma can be generalized to principal ideal domains.
Let be a principal ideal domain, and consider any . Let . Then there exist elements for which . Furthermore, is the minimal such element (under divisibility), i.e. if then .
Note that this statement is indeed a generalization of the previous statement, as the ring of integers, is a principal ideal domain.
Proof
Consider the ideal . As is a principal ideal domain, must be principle, that is it must be generated by a single element, say . Now from the definition of , there must exist such that . We now claim that .
First we prove the following simple fact: if , then . To see this, note that if , then there must be some such that . But now by definition we have .
Now from this, as , we get that . Furthermore, consider any with . We clearly have that . So indeed .
Now we shall prove minimality. Let . Then as , we have , as desired.
See also
This article is a stub. Help us out by expanding it.