Adding a Statistical Measure
A statistical measure is represented by one or more lines drawn on the chart to indicate the values of statistical quantities derived from the data (confidence intervals, percentages, percentiles, quantiles, or standard deviation).

To add a statistical measure, follow the steps below:
1. Right-click the Chart, and select 'Properties' from the context menu. This opens the 'Chart Properties' dialog box.
2. Select the Advanced tab of the 'Properties' dialog box.
3. In the 'Target Lines' panel, press the 'Add' button. This opens the 'Add Target' dialog box.
4. Select the Statistics option.
5. In the 'Field' menu, select the chart measure to which you want to add the statistical measure.
6. In the 'Computation' field, select one of the following options to compute statistics from the data: 'Confidence Interval', 'Percentages', 'Percentile', 'Quantiles', 'Standard Deviation'. (See explanations below.)

7. To modify the statistical measure, press the 'Edit' button. The following settings are available:
a. Confidence Interval: For the 'Confidence Interval' option, enter a value as a percentage.

The resulting top and bottom confidence bounds indicate the interval of values in which the “true” value is expected to be found. For example, the “true” temperature in the chart below would be expected to fall within the displayed confidence interval in 99 out of 100 such samples. (In other words, the true temperature is expected to be outside the confidence bounds purely by chance in 1 out of 100 samples.)

b. Percentages: For the 'Percentages' option, enter a value or comma-separated list of values as percentages. In the 'Of' field, specify the basis on which the percentage should be computed. You can type a fixed value or select from the following presets: 'Average', 'Minimum', 'Maximum', 'Median', 'Sum'.

For example, to display percentage lines at 70% and 90% of the Maximum, enter “70,90” in the 'Percentages' field and select the 'Maximum' option from the 'Of' field.

c. Percentiles: For the 'Percentiles' option, enter a value or comma-separated list of values as percentages.

The resulting percentile lines indicate the levels below which the specified percentages of values are found. For example, percentile lines at 70% and 90% (“70,90” in the 'Percentages' field) designate the levels, respectively, below which 70% and 90% of the data are found.

d. Quantiles: For the 'Quantiles' option, enter the number of quantiles to display.

The resulting quantile lines are evenly distributed between 0% and 100% and indicate the levels below which the specified percentage of values are found. For example, enter “4” as the 'Number of Quantiles' to generate lines designating the levels below which 25%, 50%, and 75% of the data are found. This creates four regions in the data: 0-25%, 25%-50%, 50%-75%, and 75%-100%.

e. Standard Deviation: For the 'Standard Deviation' option, enter a comma-separated list of factors. Each successive pair of factors represents, respectively, the lower and upper multipliers for the standard deviation.

For example, enter “-1,1,-2,2” in the 'Factors' field to draw lines, respectively, at 1 standard deviation below the mean, 1 standard deviation above the mean, 2 standard deviations below the mean, and 2 standard deviations above the mean.

Select the 'Sample' option to compute the sample standard deviation or select the 'Population' option to compute the population standard deviation. (The distinction between sample and population standard deviation can be found in any statistical reference.)
8. (Optional) Enable the 'Entire Chart' option to compute the statistical values based on measure data from the entire chart. Disable the 'Entire Chart' option to compute the statistical values for each sub-chart based only on measure data from the same sub-chart.
The following illustration demonstrates the effect of the 'Entire Chart' setting ('Computation' is set to 'Standard Deviation' in both cases).

9. From the 'Label' menu, select one of the following label options:

To enter a literal comma in the label, escape the comma with a backslash (e.g., “Q1\,25% below, Q2\, 50% below, Q3\,75% below”).
a. Select 'Enter a Value' to type custom labels for the statistical measures. Labels for individual lines should be separated by commas. For example, if you are generating the 4-quantile (which creates three lines), enter three labels separated by commas, e.g., “Q1: 25% below, Q2: 50% below, Q3: 75% below”.

If you enter only a single label, this label will be attached to all of the lines. This can be useful when you include customization codes in the label, as described below.
b. Select '(Target Value)' to insert the numerical value of the lines as the labels.
c. Select '(Target Formula)' to insert the name of the computation method (e.g., 'Quantile 1', 'Quantile 2', etc.) as the label, if applicable.
d. Select '(Field Name)' to insert the field name for the selected measure as the label, e.g., “Sum(Total)”.
Note: The customized label supersedes any previous selection from the 'Label' menu.
10. (Optional) To further customize the labels, press the 'Edit' button next to the 'Label' field. This opens a panel in which you can manually enter the labels. Press the green 'Apply' button when you have finished entering the labels.

To enter a literal comma in the label, escape the comma with a backslash (e.g., “Q1\,25% below, Q2\, 50% below, Q3\,75% below”).
Labels for individual lines should be separated by commas. For example, if you are generating the 4-quantile (which creates three lines), enter three labels separated by commas, e.g., “Q1: 25% below, Q2: 50% below, Q3: 75% below”.
If desired, you can add the target value, target formula, and field name into a label by inserting the corresponding codes ({0},{1},{2}) shown at the bottom of the panel. You can format the inserted values using the same syntax as Custom Tooltip. Some examples are shown below:
{1} = {0,number,$#,##0} yeilds “70% of Max = $383,485”
{1} for monthly {2} yeilds “70% of Max for monthly Sum(Total)”
11. Select a 'Line Style' and 'Line Color' in which to display the statistical lines.
12. (Optional) Press the 'Fill Band' button to open a color picker and select a set of colors for the specified bands (i.e., the regions between the statistical lines). Select one color for each band. The colors are applied to the bands from left to right; the left-most color is applied to the lowest band, and so on. When you have selected the desired colors, press the green 'Apply' button.

For more information about the features of the color-picker, see Adding Color Coding to Dimensions.
13. (Optional) Select a 'Fill Above' color and 'Fill Below' color to fill the regions of the chart above and below the maximum and minimum statistical lines, respectively.

14. Press 'OK' to close the 'Add Target' dialog box, and press 'OK' to close the 'Chart Properties' dialog box.
By default, a statistical line appears on the chart even if its value is greater than the largest data point. This may sometimes cause the data points on the chart to be compressed into a small region of the plot area, which makes the chart difficult to read. To correct this, turn off the 'Keep Element in Plot' option in the 'Plot Properties' dialog box. See Plot Property for more information.
See Also
Adding a Target Line, to place a target line or representative line on the chart.
Adding a Target Band, to demarcate a specified data region on the chart.
| << Adding a Target Band | © 1996-2013 InetSoft Technology Corporation (v11.5) | 5.11 Special Elements >> |