MATLAB: How do you separate imported data into time and amplitude

detrendhelpstudentundergraduateuniversity

%%i have imported notepad data into matlab, having difficulty seperating the data into seperate arrays. Any help?

Best Answer

If by detrend you mean that you want to remove the part of the signal that is growing linearly with time then the expression you have will not work. It will just shift your whole plot horizontally, so that it is centered on the mean value for the whole series.

Instead you can do a little linear regression to find the best straight line (y = mx + b) through the data, and then subtract that off.

To do this in MATLAB you could use:

% make regression matrix, A where
% y = A*c + error
A = [x(:) ones(size(x(:)))]
% Find least squares solution (regression) for constants c that provide best fit for
% y = c(1)*x + c(2) using Matlab \ operator
c = A\y
% remove the linear growth (detrend) the data
yDetrended = y - (c(1)*x + c(2))
% plot the result
plot(x,yDetrended)

Here is a good link to learn more about the linear regression, https://www.mathworks.com/help/matlab/data_analysis/linear-regression.html

Related Solutions

MATLAB: How do you transfer this signal into linear form

I see this is a follow up to your previous question which was answered in https://www.mathworks.com/matlabcentral/answers/506414-how-do-you-separate-imported-data-into-time-and-amplitude

The approach in the earlier post to to remove the linear trend the data was:

% make regression matrix, A where

% y = A*c + error

A = [x(:) ones(size(x(:)))]
% Find least squares solution (regression) for constants c that provide best fit for

% y = c(1)*x + c(2) using Matlab \ operator
c = A\y
% remove the linear growth (detrend) the data
yDetrended = y - (c(1)*x + c(2))
% plot the result

plot(x,yDetrended)

Looking at your new dataset signal_2.txt, it looks like the data now has a parabolic (quadratic) trend to remove.

You could modify the previous approach to instead fit a quadratic curve to the data and then subtract it off.

% make regression matrix, A where
% y = A*c + error
A = [x(:).^2 x(:) ones(size(x(:)))]
% Find least squares solution (regression) for constants c that provide best fit for
% y = c(1)*x^2 + c(2)*x + c(3) using Matlab \ operator
c = A\y
% remove the quadratic trend from the data
yDetrended = y - A*c % note A*c = (c(1)*x.^2 + c(2)*x + c(3)) 
figure
% plot the result
plot(x,yDetrended)

p.s. If you have a signal_3.txt that has a cubic trend, I think you can figure out what to do from here :)

MATLAB: Quadratic time detrending. Any help or code ??

Yes, that should work, you just need to define:

 Ynew = Y - Th;

to get the new "detrended" timeseries.

Best Answer

Related Solutions

MATLAB: How do you transfer this signal into linear form

MATLAB: Quadratic time detrending. Any help or code ??

Related Question