MATLAB®

Detrending Data

Introduction

The MATLAB^® function detrend subtracts the mean or a best-fit line (in the least-squares sense) from your data. If your data contains several data columns, each data column is detrended separately.

Removing a trend from the data enables you to focus your analysis on the fluctuations in the data about the trend. A linear trend typically indicates a systematic increase or decrease in the data. This might be caused by sensor drift, for example.

You must decide whether it makes sense to remove trend effects in the data based on the objectives of your analysis.

Back to Top

Example: Removing Linear Trends from Data

This example shows how to remove a linear trend from daily closing stock prices to emphasize the price fluctuations about the overall increase. This data is available in the Financial Toolbox™ predict_ret_data.mat file.

You can follow along with the steps in this example to perform the following tasks:

• Loading and Plotting Data

• Detrending Data and Plotting Results

Loading and Plotting Data

1. Load the sample data:

   load predict_ret_data.mat
   

   This adds the variable sdata to the workspace, which contains the daily stock prices.

2. View the contents of the column vector sdata:

   sdata
   

   The last data value is a NaN, which must be removed before detrending the data.

3. Identify and remove the NaN value from sdata:

   sdata(any(isnan(sdata),2),:) = []

   For more information about removing NaNs, see Removing NaNs from Data.

4. Plot the data:

   plot(sdata,'+')
   legend('Original Data',1);
   xlabel('Time (days)');
   ylabel('Stock Price (dollars)');

   Daily Closing Stock Prices

   

   Notice the systematic increase in the stock prices when this data was collected.

Detrending Data and Plotting Results

1. Remove a best-fit line (in the least-squares sense) from sdata and save the results to a new variable, detrend_sdata:

   detrend_sdata=detrend(sdata);

2. Plot the detrended data in a new MATLAB Figure window:

   figure
   plot(detrend_sdata,'-')
   legend('Detrended Data',2)
   xlabel('Time (days)'); 
   ylabel('Detrended Stock Price (dollars)');

   Stock Prices with the Removed Linear Trend

   

   Notice that the data is now centered about 0 and the linear drift is removed from the data.

Back to Top

Filtering Data

Differencing Data