Smooth the data and find major peaks
18 次查看(过去 30 天)
显示 更早的评论
I have attached a data which shows some periodic peaks.
Upon close observation, it can be found that the data is not continously increasing
or decreasing over a period of sinusodal curve. The data breaks after every several interval.
Can anybody help me smootheing the data and find major increasing and decreasing peaks?
I used several functions like
x2 = smooth(x1);
x2 = smoothdata(x1,'gaussian',20);
x2 = smooth(x1,'loess');
There are still errors and I cannot observe peaks.
Can anyone please help.
1 个评论
Adam
2018-12-19
If you have the SIgnal Processing Toolbox then
doc findpeaks
can help find the peaks.
I don't know what you mean by 'there are still errors' though. There are numerous ways to smooth data, it depends on the data and what is noise and what is true data that should be retained. I have sometimes used Savitzky–Golay filters for this, but gaussian smoothing or even mean smoothing can work depending how aggressive you need the smoothing to be.
采纳的回答
Luna
2018-12-20
编辑:Luna
2018-12-20
findpeaks is built-in function of Matlab and it already works for detecting peaks and locations. Try below:
load('matlab.mat');
y1 = GGG1_qzss;
x1 = (1:numel(y1))'; % creates a time series for 1 seconds each you can use different time samples also
figure;
plot(x1,y1,'r');
hold on;
MPH = 8; % minimum peak height you can change it if you want
MINW = 100; % minimum peak width
[pks,locs] = findpeaks(y1,x1,'MinPeakHeight',MPH,'MinPeakWidth',MINW);
% pks: peak values
% locs: location of peaks in array
plot(x1(locs),pks,'bo');
更多回答(1 个)
Sarah Crimi
2018-12-19
编辑:madhan ravi
2018-12-19
There is an open source function called peakdet.m. Using this function, you have to set the delta which determines what is a peak. For instance when I used this,
delta1=0.000005;
[max_m, min_m]=peakdet(m, delta1);
The link the open source file is below:
0 个评论
另请参阅
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!