1

I am new to this forum, so excuse me if I don't get my question right at the first start. I have researched in the forum to find an answer to my problem but I haven't found a proper solution yet. I have a data set of k time series and provided that I have put the one as reference, I am trying to compare the other time series with the reference in order to figure out which of them are similar to the reference. I am using methods such as LCSS, ERP, EDR, DTW but eventually I have some problems. The time series are not of equal length, so every method doesn't give me all the similar time-series that it should so because of the different length. I am looking for a way to transform time series of different length into a unique length. How can I do this in R?

Nikos
  • 11
  • 2
  • First of all, it would be very helpful if you could provide a [reproducible example](https://stackoverflow.com/q/5963269/1861328) which illustrates your problem. Without sample data set, it is really difficult to answer your question. Use [`dput(your_data)` ](https://stackoverflow.com/a/5963610/1861328) to copy your data to the forum, or prepare dummy data set which reproduce the structure of your data. Include your code. Help people to understand your problem. – utubun Jun 21 '18 at 18:53
  • You should say if you want the shorter to grow or the bigger to be clipped and how in any of those cases. – Rui Barradas Jun 21 '18 at 18:53
  • Look at the `approx` function. – Alexis Jun 22 '18 at 23:49

1 Answers1

1

You could check for the implementations Piecewise Aggregate Approximation (PAA). There have been a lot of improvements. You will find a lot of implementations of PAA in python, not so sure about R. However the underlying algorithm is pretty straightforward. Here are some reference links, [1], [2]

Deven Mistry
  • 55
  • 5