Absolute convergence

Absolute convergence (Redirected from Absolute convergence theorem) Jump to navigation Jump to search This article includes a list of general references, mas faltam citações em linha correspondentes suficientes. Ajude a melhorar este artigo introduzindo citações mais precisas. (Fevereiro 2013) (Saiba como e quando remover esta mensagem de modelo) Na matemática, an infinite series of numbers is said to converge absolutely (or to be absolutely convergent) if the sum of the absolute values of the summands is finite. Mais precisamente, a real or complex series {soma de estilo de texto de estilo de exibição _{n=0}^{infty }uma_{n}} is said to converge absolutely if {soma de estilo de texto de estilo de exibição _{n=0}^{infty }deixei|uma_{n}certo|=L} for some real number {displaystyle textstyle L.} De forma similar, an improper integral of a function, {displaystyle textstyle int _{0}^{infty }f(x),dx,} is said to converge absolutely if the integral of the absolute value of the integrand is finite—that is, E se {displaystyle textstyle int _{0}^{infty }|f(x)|dx=L.} Absolute convergence is important for the study of infinite series because its definition is strong enough to have properties of finite sums that not all convergent series possess - a convergent series that is not absolutely convergent is called conditionally convergent, while absolutely convergent series behave "nicely". Por exemplo, rearrangements do not change the value of the sum. This is not true for conditionally convergent series: The alternating harmonic series {estilo de texto 1-{fratura {1}{2}}+{fratura {1}{3}}-{fratura {1}{4}}+{fratura {1}{5}}-{fratura {1}{6}}+cdots } converge para {estilo de exibição ln 2,} while its rearrangement {estilo de texto 1+{fratura {1}{3}}-{fratura {1}{2}}+{fratura {1}{5}}+{fratura {1}{7}}-{fratura {1}{4}}+cdots } (in which the repeating pattern of signs is two positive terms followed by one negative term) converge para {estilo de texto {fratura {3}{2}}ln 2.} Conteúdo 1 Fundo 1.1 Explicação 2 Definition for real and complex numbers 3 Sums of more general elements 3.1 In topological vector spaces 4 Relation to convergence 4.1 Proof that any absolutely convergent series of complex numbers is convergent 4.1.1 Alternative proof using the Cauchy criterion and triangle inequality 4.2 Proof that any absolutely convergent series in a Banach space is convergent 5 Rearrangements and unconditional convergence 5.1 Real and complex numbers 5.2 Series with coefficients in more general space 5.3 Prova do teorema 6 Products of series 7 Absolute convergence over sets 8 Absolute convergence of integrals 9 Veja também 10 Notas 11 Referências 11.1 Works cited 11.2 General references Background In finite sums, the order in which terms are added does not matter. 1 + 2 + 3 is the same as 3 + 2 + 1. No entanto, this is not true when adding infinitely many numbers, and wrongly assuming that it is true can lead to apparent paradoxes. One classic example is the alternating sum {displaystyle S=1-1+1-1+1-1...} whose terms alternate between +1 e -1. What is the value of S? One way to evaluate S is to group the first and second term, the third and fourth, e assim por diante: {estilo de exibição S_{1}=(1-1)+(1-1)+(1-1)....=0+0+0...=0} But another way to evaluate S is to leave the first term alone and group the second and third term, then the fourth and fifth term, e assim por diante: {estilo de exibição S_{2}=1+(-1+1)+(-1+1)+(-1+1)....=1+0+0+0...=1} This leads to an apparent paradox: does {displaystyle S=0} ou {displaystyle S=1} ?

The answer is that because S is not absolutely convergent, rearranging its terms changes the value of the sum. This means {estilo de exibição S_{1}} e {estilo de exibição S_{2}} are not equal. Na verdade, the series {estilo de exibição 1-1+1-1+...} does not converge, so S does not have a value to find in the first place. A series that is absolutely convergent does not have this problem: rearranging its terms does not change the value of the sum.

Explanation This is an example of a mathematical sleight of hand. If the terms of S are rearranged in such a way that every term remains in its original position, one finds that S is either the infinite series {displaystyle S=1-1+1-1+...+1-1+1-1} or with equal possibility, este {displaystyle S=1-1+1-1+...+1-1+1} Evaluating S as before, by grouping every -1 with the +1 preceding it or by grouping every +1 except the first with the -1 preceding it, gives in the first case: {estilo de exibição S_{1}=(1-1)+....+(1-1)=0+....+0=0} {estilo de exibição S_{2}=1+(-1+1)+....+(-1+1)-1=1+0+....+0-1=1-1=0} e no segundo caso: {estilo de exibição S_{1}=(1-1)+....+(1-1)+1=0+....+0+1=1} {estilo de exibição S_{2}=1+(-1+1)+....+(-1+1)=1+0+...+0=1} This reveals the trick: the definition of S was interpreted as defining its last term as negative when evaluating {estilo de exibição S_{1}=0} but positive when evaluating {estilo de exibição S_{2}=1} when in fact the definition of S didn't define (and the rearrangement was independent of) either option.

Definition for real and complex numbers A sum of real numbers or complex numbers {textstyle sum _{n=0}^{infty }uma_{n}} is absolutely convergent if the sum of the absolute values of the terms {textstyle sum _{n=0}^{infty }|uma_{n}|} converge.

Sums of more general elements The same definition can be used for series {textstyle sum _{n=0}^{infty }uma_{n}} whose terms {estilo de exibição a_{n}} are not numbers but rather elements of an arbitrary abelian topological group. Nesse caso, instead of using the absolute value, the definition requires the group to have a norm, which is a positive real-valued function {estilo de texto |cdot |:Gto mathbb {R} _{+}} on an abelian group {estilo de exibição G} (written additively, with identity element 0) de tal modo que: The norm of the identity element of {estilo de exibição G} é zero: {estilo de exibição |0|=0.} Para cada {displaystyle xin G,} {estilo de exibição |x|=0} implica {displaystyle x=0.} Para cada {displaystyle xin G,} {estilo de exibição |-x|=|x|.} Para cada {estilo de exibição x,yin G,} {estilo de exibição |x+y|leq |x|+|y|.} Nesse caso, a função {estilo de exibição d(x,y)=|x-y|} induces the structure of a metric space (a type of topology) sobre {displaystyle G.} Então, uma {estilo de exibição G} -valued series is absolutely convergent if {textstyle sum _{n=0}^{infty }|uma_{n}|0,} there exists {displaystyle N} such {textstyle left|sum _{i=m}^{n}left|a_{i}right|right| =sum _{i=m}^{n}|a_{i}|mgeq N.} But {textstyle {big |}sum _{i=m}^{n}a_{i}{big |}leq _{i=m}^{n}|a_{i}|,} so {textstyle left|sum _{i=m}^{n}a_{i}right|mgeq N,} exactly {textstyle sum a_{i}.} above result easily generalized {displaystyle (X,|,cdot ,|).} {textstyle sum x_{n}} {displaystyle X.} As {textstyle sum _{k=1}^{n}|x_{k}|} numbers, {displaystyle varepsilon >0} large enough natural {displaystyle m>n} holds: {displaystyle left|sum _{k=1}^{m}|x_{k}|-sum _{k=1}^{n}|x_{k}|right| =sum _{k=n+1}^{m}|x_{k}|kappa _{varepsilon }&quad sum _{n=N}^{infty }|uma_{n}|<{tfrac {varepsilon }{2}}\{text{ for all }}N>lambda _{varepsilon }&quad left|soma _{n=1}^{N}uma_{n}-Aright|<{tfrac {varepsilon }{2}}end{aligned}}} Let {displaystyle {begin{aligned}N_{varepsilon }&=max left{kappa _{varepsilon },lambda _{varepsilon }right}\M_{sigma ,varepsilon }&=max left{sigma ^{-1}left(left{1,ldots ,N_{varepsilon }right}right)right}end{aligned}}} where {displaystyle sigma ^{-1}left(left{1,ldots ,N_{varepsilon }right}right)=left{sigma ^{-1}(1),ldots ,sigma ^{-1}left(N_{varepsilon }right)right}} so that {displaystyle M_{sigma ,varepsilon }} is the smallest natural number such that the list {displaystyle a_{sigma (0)},ldots ,a_{sigma left(M_{sigma ,varepsilon }right)}} includes all of the terms {displaystyle a_{0},ldots ,a_{N_{varepsilon }}} (and possibly others). Finally for any integer {displaystyle N>M_{sigma ,varepsilon }} deixar {estilo de exibição {começar{alinhado}EU_{sigma ,varepsilon }&=left{1,ldots ,Nright}setminus sigma ^{-1}deixei(deixei{1,ldots ,N_{varepsilon }certo}certo)\S_{sigma ,varepsilon }&=min sigma left(EU_{sigma ,varepsilon }certo)=min left{sigma (k) : kin I_{sigma ,varepsilon }certo}\EU_{sigma ,varepsilon }&=max sigma left(EU_{sigma ,varepsilon }certo)=max left{sigma (k) : kin I_{sigma ,varepsilon }certo}\fim{alinhado}}} de modo a {estilo de exibição {começar{alinhado}deixei|soma _{iin I_{sigma ,varepsilon }}uma_{sigma (eu)}certo|&leq sum _{iin I_{sigma ,varepsilon }}deixei|uma_{sigma (eu)}certo|\&leq sum _{j=S_{sigma ,varepsilon }}^{EU_{sigma ,varepsilon }}deixei|uma_{j}certo|&&{texto{ desde }}EU_{sigma ,varepsilon }subseteq left{S_{sigma ,varepsilon },S_{sigma ,varepsilon }+1,ldots ,EU_{sigma ,varepsilon }certo}\&leq sum _{j=N_{varepsilon }+1}^{infty }deixei|uma_{j}certo|&&{texto{ desde }}S_{sigma ,varepsilon }geq N_{varepsilon }+1\&<{frac {varepsilon }{2}}end{aligned}}} and thus {displaystyle {begin{aligned}left|sum _{i=1}^{N}a_{sigma (i)}-Aright|&=left|sum _{iin sigma ^{-1}left({1,dots ,N_{varepsilon }}right)}a_{sigma (i)}-A+sum _{iin I_{sigma ,varepsilon }}a_{sigma (i)}right|\&leq left|sum _{j=1}^{N_{varepsilon }}a_{j}-Aright|+left|sum _{iin I_{sigma ,varepsilon }}a_{sigma (i)}right|\&0,{texto{ existe }}M_{sigma ,varepsilon },{texto{ para todos }}N>M_{sigma ,varepsilon }quad left|soma _{i=1}^{N}uma_{sigma (eu)}-Aright|

Se você quiser conhecer outros artigos semelhantes a Absolute convergence você pode visitar a categoria Integral calculus.