Given that SUM_{n=1}^{infinity}(u_n) is a divergent series of positive terms, show that SUM_{n=1}^{infinity}(u_n)^p diverges for 0 < p < 1.
That is, \[\sum_{n=1}^{\infty}u_n\] diverges is given. I need to show that \[\sum_{n=1}^{\infty}u_n^p\] also diverges for 0<p<1.
"comparison test" i believe is the term of art
since what jamesj said.
\[u_n\] are positive, and \[u_n^p > u_n\] is only true if \[u_n < 1\] which isn't something one can assume...
yes sorry. It's more subtle than that, one sec.
hold on. if after some n \[u_n>1\] it diverges for sure right?
in which case you are done
yeah, but we are not given that the \[{u_n}\] form an increasing sequence...
Right, thank-you. If it's not true that there is some N such that \[ n > N \implies |u_n| < 1 \] then u_n^p stays greater than 1 also and hence diverges. But if it is true that the tail of (u_n) goes under 1, then you can use the fact that \[ u_n^p > u_n \] for those terms.
Maybe they do as a consequence of the given diverging sum?
Yeah James I think that you are right...
only the tail determines convergence or divergence
In other words: \[ u_n \geq 1 \implies u_n^p \geq 1 \] \[ u_n < 1 \implies u_n^p > u_n \] hence we may bound every term of \( u_n^p \) by \[ u_n^p \geq \min(1,u_n) \] and the sum of those terms has to diverge.
it does say the are positive terms. if there does not exist and N for which all n>N you have \[u_n<1\] then it diverges because the terms do not even go to zero. so we can assume then that there does exist such an N and use what jamesj said above, i.e. comparison test
more eloquently put by jamesj
Very good, thanks both of you!
and I promise to drink at least one cup of coffee before attempting such questions in the future.
Join our real-time social learning platform and learn together with your friends!