Displaying a Trend Line in a graph visualization

You can display a trend lines in a graph visualization for each metric or for a specific metric only. A trend line represents trends in your data. You can extend a trend line beyond your existing data to forecast or predict future values.

Trend lines are not available for stacked or clustered graphs for which there is more than one metric series displayed in the graph, or percent graphs. At least one axis must contain numeric values.

1. Click on the graph visualization you want to modify.

2. Right-click the numeric axis to use as a basis for the trend line. If the axis contains more than one metric, choose the metric name. Then, choose Enable Trendline.

3. Once you enable a trend line, you can edit it. Right-click the trend line and select Format.

4. Select the line type from the first drop-down in Line. Next, select the color of the line from the second drop-down.

5. You can select several different types of trend lines from the Model drop-down list.

   Select Best Fit to display a trend line that best fits the data. This is the trend line with the highest r-squared value.

   or

   Select Linear to display a linear trend line, which is a best-fit straight line that is used with simple linear data sets. You data is linear if the pattern in its data points resembles a line. A linear trend line represents data that is increasing or decreasing at a steady rate.

   or

   Select Logarithmic to display a logarithmic trend line, which is a best-fit curved line that is most useful when the rate of change in the data increases or decreases quickly and then levels out. This option is available if the graph values are positive.

   or

   Select Exponential to display a exponential trend line, which is a curved line that is most useful when data values rise or fall at increasingly higher rates.

   or

   Select Power to display a power trend line, which is a curved line that is best used with data sets that compare measurements that increase at a specific rate.

   or

   Select Polynomial to display a polynomial trend line, which is a curved line that is used when data fluctuates. This trend line allows you to analyze gains and losses over a large data set. Then, select the degree of the polynomial from the Degree drop-down. The degree of the polynomial corresponds to the number of upward and downward fluctuations in the data. Your data fluctuates if the pattern in its data points resembles a line with hills and valleys. A polynomial trend line with a degree of two contains up to one hill or valley, a degree of three contains up to two hills or valleys, etc.

6. Use the options in the Level drop-down list to display a single trend line across each graph in the visualization or use the same trend line across each row, column, or individual graph. The availability of the Level option depends on the layout of your graph.

   Select Per cell to display a separate trend line for each chart in the visualization.

   or

   Select Per row to display a separate trend line for each row of charts in the visualization.

   or

   Select Per column to display a separate trend line for each column of charts in the visualization.

   or

   Select Entire table to have each chart share the same trend line.

7. A separate trend line appears for each attribute in the Break By area of the Editor panel. Clear the Include Break By Attribute checkbox to display a single line for each attribute.

   The Include Break By Attribute checkbox is available for graphs other than scatter or bubble, and if the graph has metrics on only one axis.

8. You can enable forecasting, which displays the trend line past the available data. If the graph contains metrics on only one axis, go to step 9. If the graph contains metrics on both axes, go to step 11.

9. Select the Enable forecasting checkbox and enter the number of periods for displaying data before the start and end points of the graph in Forward and Backward.

10. If the graph uses time-based attributes, such as Week, Month, Quarter, Year and Decade, select the period from the Seasons drop-down list. Go to step 16.

11. Select the metric to display the trend line for from the Forecasting drop-down list.

12. In the pop-up dialog, select the Enable Forecasting checkbox.

13. In the Min field, enter the minimum value at which to display the trend line. This value must be less than the minimum value available in the data.

14. In the Max field, enter the maximum value at which to display the trend line. This value must be greater than the maximum value available in the data.

15. Click OK.

16. To remove a trend line, right-click the trend line and select Disable trend line.