Hi I need to find periodic outliners in time series data with python. I searched a long time but all the outliner-finding techniques I found would only find one outliner and not recurrent ones. I would realy appreciate if you could give me a hint how I could do this.
Asked
Active
Viewed 131 times
0
-
Does this answer your question? [Pythonic way of detecting outliers in one dimensional observation data](https://stackoverflow.com/questions/22354094/pythonic-way-of-detecting-outliers-in-one-dimensional-observation-data) – Xukrao Feb 01 '20 at 17:03
-
Is the period of your time series a known parameter? Is your time series sampled at fixed intervals? If so, could you simply wrap all observations and take the mean at each "phase" point? – NicholasM Feb 01 '20 at 17:28
-
By the way, other users will be able to help much better if you could provide a simple, copy-pastable data set, with examples of points that you *would* and *would not* consider outliers. I'm suggesting code that uses numpy and can be pasted and run in a Python REPL. – NicholasM Feb 01 '20 at 17:29