Filter time-variant data and reduce its length

Hi experts,

I have a time-variant test data. The 1st column is time, and the 2nd column is data. There is a serious fluctuation of it. I want to filter the data and at the same time want to reduce its length.

Could somebody guide me how to do this? Thank you!

Answer 1

[Ameer Hamza]

It seems like you are trying to plot step-response of a system. Try something like this

data = load('data.csv');

t = data(1:100,1);
y = data(1:100,2);

y = smoothdata(y, 'sgolay', 20);

plot(t, y)

It reduces the length by taking the first 100 elements and smoothening the signal.

[zhi liu]

I want to get the full-response but a shorter data is needed. Is there suitable method? Thank you!

[Ameer Hamza]

Yes, The system reaches its steady-state after a few seconds. The rest of the data is redundant. This gives the full response of the system.

Answer 2

[Thiago Henrique Gomes Lobato]

Take a look at medfilt1 and/or movmean. They sould help you reduce the fluctuations. To reduce the data length you just index the time you want:

idx1 = 100;
idx2 = 900;
Data = Data(idx1:idx2,:); % Data = Length x 2 matrix

In this example your data will go only from meausurement point 100 to measurement point 900.