Time series index

blahblahbla

What does it mean to index a time series?
Ive been doing some reading and one common feature is that a lot of research gone into indexing using Euclidian distance.

Fremen

From the information you've provided, it's not entirely clear. If you could provide more context, that would be helpful.

Usually, an "index set" is just a bunch of "things". Really- no more precise than that. Within the context of probability, to index a stochastic process means to assign a random variable to each "thing" in the index set.

Common examples of index sets are:

The natural numbers (giving rise to a discrete set of variables x(1), x(2), ...)

The real numbers (giving rise to X(t) for any real number)

You can also index a stochastic process using more general objects, like the Euclidean plane - for any point (a,b), you have a random variable X(a,b).

Part of the difficulty in answering this is that I'm not sure what the formal definition of a time series is. The "time" in time series kind of suggests there should be a natural total ordering on the set.

blahblahbla

Sorry about that.

Its with regard to this
http://alumni.cs.ucr.edu/~mvlachos/pubs/kdd03.pdf

It talks about DTW and Euclidian and Im just wondering is it the case where people say trying to index time series they mean comparing different time series to one another?
For example
For 3 time series a,b,c
a is closer to b,b is closer to c and a is not as close to c as b is