Time series analysis

[lje032]

I have temperature versus time data for 8 subjects in a research study who were exposed to cooling for 30mins (~150 data point). The temperature calls as you would expect but it also looks like the pattern of fluctuation seen during baseline changes with cooling (smaller and fewer) Can anyone help me with where to start to quantify these changes (?amplitude and frequency) in a simple way. I am a PG clinical student, I have tried reading about time series analysis and covered a module with the OU 2years ago but just getting more and more confused. Any suggestions gratefully received.

 

[Overlow]

Hi guy,

If I well understand your need, you have just filter your timeserie.
Maybe you know, in most of cases a timeserie can be decomposed in three components: trend, seasonality and residuals.

Moving average is a first good way to vizualize your serie without seasonality.
You'll have the global patterns of your serie (temperature increase or decrease).

I hope to help you.

Cheers.