Geek Answers Handbook

Smoothing time series data without smoothing peak values ​​in R

I have a 3-day series of daily data (data is recorded every 5 minutes). The data is pretty noisy.
I have already tried some MA methods. They work fine, and the resulting curve is pretty smooth, but the problem is that the peaks are almost smooth.

So my question is:

Is there a way to get rid of all this noise on the chart, but keep the peak values?

I also read something about Kalman-Filtering, but I'm not sure how this works, and if that works for my problem.

I tried the following code:

smooth <- rollapply(PCM4 [,3], width=10, FUN=mean, align = "center", fill=NA)

I also tried several different input values ​​for the window width, which made the resulting data smoother, but also reduced the peak values, which are not what I want.

data set:

DateTime            h     v     Q      T
2014-12-18 11:45:00 0.112 0.515 17.141 15.4
2014-12-18 11:50:00 0.113 0.511 17.007 15.5
2014-12-18 11:55:00 0.114 0.518 17.480 15.5

unsweetened plot:

unsmoothed plot

( = 10):

enter image description here

, , , 250 / 500 /. , , .

: ?

+4
1

, ?

, , . , ( "" ) . , ( ). .

Kalman , . "" KF , , . ( , ), KF .

, , - http://www.lifl.fr/~casiez/1euro/ "1 ". , ( ), , , . , , . , .

+2

More articles:


All Articles