How to smooth and normalize noisy data

16 views (last 30 days)
I plotted sumrate against number of iterations but my data is very noisy. I need a smooth PDF, how can I smooth and normalize the attached noisy data. Thanks.
  2 Comments
Felix Obite
Felix Obite on 18 Sep 2022
i need to improve on this, have tried several smoothing function.

Sign in to comment.

Accepted Answer

Star Strider
Star Strider on 18 Sep 2022
I have no idea what result you want.
One option is the filloutliers function (or similar functions) and another is the Signal Processing Toolbox sgolayfilt function to do broadband filtering.
  2 Comments
Felix Obite
Felix Obite on 18 Sep 2022
Edited: Felix Obite on 18 Sep 2022
hi, i tried sgolayfilt function, but got error message. so I used the following function:
pro_position = [451 461 913 271];
smooth_rate = 0.01; % other pdfs with less smoothing
h = figure;
h.Position = pro_position;
plot(i,smooth(x, smooth_rate))
And this is the best I could get. I need to improve the convergence plot.
Star Strider
Star Strider on 18 Sep 2022
It would be interesting to see your sgolayfilt call. I usually use a degree 3 polynomial, begin with a frame length that is about 5% of the length of the vector, and then adjust the framelength until I get the result I want.
There are several smooth functions. I have no idea which one you are referring to.

Sign in to comment.

More Answers (1)

KALYAN ACHARJYA
KALYAN ACHARJYA on 18 Sep 2022
Please check the smoothdata Matlab function
it can smooth the data in number of ways, note that the trade-off between the actual data points and the averaged data points should be within the acceptable level.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!