Learn more about System Identification Toolbox

detrend - Subtract offset or trend from data signals

Syntax

data_d = detrend(data)
data_d = detrend(data,Type)
[data_d,T] = detrend(data,Type)
data_d = detrend(data,1,brkp)

Description

data_d = detrend(data) subtracts the mean value from each time-domain or time-series signal data. data_d and data are iddata objects.

data_d = detrend(data,Type) subtracts a mean value from each signal when Type = 0, a linear trend (least-squares fit) when Type = 1, or a trend specified by a TrendInfo object when Type = T.

[data_d,T] = detrend(data,Type) stores the trend information as a TrendInfo object T.

data_d = detrend(data,1,brkp) subtracts a piecewise linear trend at one or more breakpoints brkp. brkp is a data index where discontinuities between successive linear trends occur. When brkp contains breakpoints that match the time vector, detrend subtracts a continuous piecewise linear trend. You cannot store piecewise linear trend information as an output argument.

Examples

Subtract mean values from input and output signals and store the trend information:

% Load SISO data containing vectors u2 and y2.
load dryer2
% Create data object with sampling interval of 0.08 sec.
data=iddata(y2,u2,0.08)
% Plot data on a time plot. Data has a nonzero mean.
plot(data)
% Remove the mean from the data.
[data_d,T] = detrend(data,0)
% Plot detrended data on the same plot.
hold on
plot(data_d)

Remove specified offset from input and output signals:

% Load SISO data containing vectors u2 and y2.
load dryer2
% Create data object with sampling time of 0.08 sec.
data=iddata(y2,u2,0.08)
plot(data)
% Create a TrendInfo object for storing offsets and trends.
T = getTrend(data)
% Assign offset values to the TrendInfo object.
T.InputOffset=5;
T.OutputOffset=5;
% Subtract offset from the data.
data_d = detrend(data,T)
% Plot detrended data on the same plot.
hold on
plot(data_d)

Subtract several linear trends that connect at three breakpoints [30 60 90]:

data = detrend(data,1,[30 60 90]);
% [30 60 90] are data indexes where breakpoints occur.

Subtract a mean value from the input signal and a V-shaped trend from the output signal, such that the V peak occurs at the breakpoint value of 119:

zd1 = z(:,:,[]); zd2 = z(:,[],:);
zd1(:,1,[]) = detrend(z(:,1,[]),1,119);
zd2(:,[],1) = detrend(z(:,[],1));
zd = [zd1,zd2];

See Also

getTrend | retrend | TrendInfo

How To

• Handling Offsets and Trends in Data

delayest

diff

© 1984-2009- The MathWorks, Inc.    -   Site Help   -   Patents   -   Trademarks   -   Privacy Policy   -   Preventing Piracy   -   RSS