Thread: Normalizing unknown time series data

Hi

I was wondering if it is possible to normalize a time series data such that their sum equals to 1, subject to the following constraints;

1. We do not know the whole series in advance.
2. We know the number of elements in the series, let it be n.
3. We need to normalize them such that their sum equals to 1.

Example;
Let the data be : x1, x2, x3 ........., xn
We do not know the value of x1, x2... etc
We want to normalize it such that their sum equal to 1.

Thanks.

Cheers.

Originally Posted by Quest

Hi

I was wondering if it is possible to normalize a time series data such that their sum equals to 1, subject to the following constraints;

1. We do not know the whole series in advance.
2. We know the number of elements in the series, let it be n.
3. We need to normalize them such that their sum equals to 1.

Example;
Let the data be : x1, x2, x3 ........., xn
We do not know the value of x1, x2... etc
We want to normalize it such that their sum equal to 1.

Thanks.

Cheers.

It appears you want non-causal processing, which is impossible unless you either wait until all the data is in or know something more about the process.

CB