Convergence of Infinite Product

Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
5 replies · 3K views
holomorphic
Messages
91
Reaction score
0

Homework Statement


If [tex]\sum a_{j}[/tex] converges absolutely, and [tex]a_{j}\neq -1[/tex] for all j, then show [tex]\prod _{j=1} ^{\infty} (1+a_{j})\neq 0[/tex]. Hint: Consider [tex]b_{j}[/tex] such that [tex](1+b_{j})(1+a_{j})=1[/tex]. Show that [tex]\sum _{j=1} ^{\infty} b_{j}[/tex] converges absolutely, and consider [tex]\prod _{j=1} ^{\infty} (1+a_{j}) \bullet \prod _{j=1} ^{\infty} (1+b_{j})[/tex]

Homework Equations


The Attempt at a Solution


Taking the hint gives [tex]b_{j} = \frac{1}{1 + a_{j}} - 1[/tex], but I am not really sure how to show [tex]\sum b_{j}[/tex] converges absolutely. I tried writing down inequalities I know, e.g. [tex]\left|a_{j} + 1 \right| \leq \left|a_{j}\right| + 1[/tex], and manipulating them to show that [tex]\sum \left|b_{j}\right| \leq \sum\left|a_{j}\right|[/tex]... but it's not working. I also tried to write [tex]\sum \left|b_{j}\right|[/tex] as a fraction with the product [tex]\prod (1+a_{j})[/tex] in the denominator, but the formula for the numerator turned out not to be so easy to write.

Any suggestions would be appreciated :)
 
Last edited:
Physics news on Phys.org
bj=1/(1+aj)-1=(-aj)/(1+aj). lim aj->0. So for large enough j, |aj|<(1/2). Isn't that enough to show bj converges absolutely if aj does?
 
Dick said:
bj=1/(1+aj)-1=(-aj)/(1+aj). lim aj->0. So for large enough j, |aj|<(1/2). Isn't that enough to show bj converges absolutely if aj does?

I guess I don't understand why that's enough to show bj converges absolutely.

Ohh wait... so [tex]\left| b_{n} \right| \leq \left| a_{n} \right|[/tex] when n>=N for some N and an converges absolutely, therefore bn converges absolutely. Right?
 
holomorphic said:
I guess I don't understand why that's enough to show bj converges absolutely.

Ohh wait... so [tex]\left| b_{n} \right| \leq \left| a_{n} \right|[/tex] when n>=N for some N and an converges absolutely, therefore bn converges absolutely. Right?

I wouldn't say |b_n|<=|a_n|. a_n isn't necessarily positive. So |1+a_n| isn't greater than one. But you can show |b_n|<=C*|a_n| for some constant C.
 
Dick said:
I wouldn't say |b_n|<=|a_n|. a_n isn't necessarily positive. So |1+a_n| isn't greater than one. But you can show |b_n|<=C*|a_n| for some constant C.

So, supposing |a_n| < 1/2, then [tex]\left| b_{n} \right| = \frac{\left| a_{n} \right|}{\left| 1 + a_{n} \right|} < 1[/tex], so that |b_n| < 2*|a_n| ?

If this is right, then thanks very much for your help.
 
holomorphic said:
So, supposing |a_n| < 1/2, then [tex]\left| b_{n} \right| = \frac{\left| a_{n} \right|}{\left| 1 + a_{n} \right|} < 1[/tex], so that |b_n| < 2*|a_n| ?

If this is right, then thanks very much for your help.

Right.