Show TOC

 Adding, Changing, and Removing Trend LinesLocate this document in the navigation structure

Use

You can visualize the trend of a data series using a trend line for a number of chart types. You can also extend the trend line past the data series in the chart in order to forecast future values and derive past values. There are the following types of trend lines:

  • Linear: A linear trend line is an optimized straight line that is suitable for linear datasets. A dataset is linear if the pattern of data points resembles a line. A linear trend line normally displays a constant increase or decrease in values.
  • Logarithmic: A logarithmic trend line is an optimized curve that is ideal if the rate of changes to the data first increases or decreases sharply and then stays nearly the same. A logarithmic trend line can use negative and/or positive values.
  • Potential: A potential trend line is a curve. It should be used for datasets that compare measurements that increase with a certain rate (for example, the acceleration of a racing car in 1 second intervals). You cannot add a potential trend line if your data contains null or negative values.
  • Exponential: An exponential trend line is a curve that is suitable for data values that increase or decrease with an increasing tendency. You cannot add an exponential trend line if your data contains null or negative values.
  • Average: The average value as calculated from multiple values. If there is only a small amount of data, large deviations in either direction can have a significant influence on the average.
  • Moving average: Smoothes out data fluctuations and can therefore show a trend or pattern more clearly. The moving average consists of a series of averages that are calculated from elements of the data series. A moving average uses a certain number of data points, computes their mean value and then uses this mean value as a point in the curve.
  • Gaussian or normal distribution: A method of statistics that records the frequency with which values are distributed within a value range. Normal distribution is particularly useful with histograms.

You can change all the properties of a trend line that you can define when you add the trend line. You can delete a trend line that you do not need.

Prerequisites

The chart is a bar, area, line, portfolio, profile, point, or column diagram or a histogram.

Adding Trend Lines
  1. In the property Data Series, select the data series that you want to add to the trend line (for example, Data Series 1).
  2. Use the secondary mouse button to click on the data series.
  3. In the context menu choose Add Analysis Function.

    A new element 'trend line' is added beneath the node of the data series.

  4. Select this element to display the properties.
  5. Choose the mathematical calculation you require in the Function Type field.
  6. Format the line as required (see Formatting Lines ).
  7. You can also color the areas above and below the trend line. To do so, choose the settings you require in the properties Upper Area and Lower Area (see Formatting Areas ).
Removing Trend Lines
  1. Use the secondary mouse button to click on the trend line you want to remove.
  2. Choose Delete Analysis Function.