Suppose I am told I have N numbers and that the sum of the square of those numbers is S. Can anything be said about the sum of those original N numbers?


Cauchy-Schwarz gives $$\sum_{n=1}^N |x_n| \le \sqrt{N} \sqrt{\sum_{n=1}^N x_n^2}.$$

For a lower bound, we have $$\left(\sum_{n=1}^N |x_n|\right)^2 \ge \sum_{n=1}^N x_n^2.$$

So, $\sqrt{S} \le \sum_{n=1}^N |x_n| \le \sqrt{NS}$.

Without further information, these bounds cannot be improved. If all of the numbers are the same, then $\sum_{n=1}^N |x_n| = \sqrt{NS}$. If all of the numbers are zero except one of them, then $\sum_{n=1}^N |x_n| = \sqrt{S}$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.