Box plot chart allows you to display differences between populations/groups of data without making any assumptions of the underlying statistical distribution – they are non-parametric. The spacing between the different parts of the box helps indicate the degree of dispersion (spread) and skewness in the data, and identify outliers. Box plot is a convenient way of graphically depicting groups of numerical data through their quartiles having lines extending vertically from the boxes (whiskers) indicating variability outside the upper and lower quartiles.

Open

To access a Bar & Line chart, go to Analytics > Data Analyzer > Box Plot. The first tab Chart contains the result chart. The second tab contains the data that is shown in the chart.

Specify Box Plot Details

You can specify the Box plot details by doing the following:

1. Enter the name of the data series in the Data Label field (max 60 characters).

2. Choose the Data Source from the drop-down list. It can be any Pricefx data source – a Datamart, Data Feed, Simulation Datamart, Rollup or Optimization Model (DMT tables organized by Models to which they belong).

3. Select a Currency.

4. You can set up the Generic filter (all dimensions, dates and values):

   1. Click Add Generic Filter, the Generic Filter dialog will open. 

   2. Specify the filter criteria. You can combine individual rules with groups of rules as clauses and subclauses.

   3. Click Apply.

5. Product filter allows you to select products from the drop-down list or filter by all attributes from the product master data. You can use the quick filters under the table header, set up an advanced filter or select an attribute value in the table, which selects all products with the same attribute value. Attributes that are not available for filtering are grayed out.

6. Customer filter allows you to select customers from the drop-down list or filter by all attributes from the customer master data. You can use the quick filters under the table header, set up an advanced filter or select an attribute value in the table, which selects all customers with the same attribute value. Attributes that are not available for filtering are grayed out. Only the filtered data will be used in the analysis.

7. You can set up a Dimension filter:

   1. Select the filter dimension from the drop down list.

   2. Select the values to filter by.

   3. Add more dimensions if required. In all subsequent filters you will only be able to select values complying with all previous filters.

8. Define the axis X and axis Y:

   1. Select the appropriate measurements for the axis X and axis Y from the drop down list.

   2. Specify how the value for axis Y will be calculated. For details see Data Calculation Options.

9. Define the Whisker:

   • No Whiskers – Shows only the box.

   • Min-Max – Lower whisker shows the minimum of all data, upper whisker shows the maximum.

   • 10-90th – 01-99th – Lower whisker shows the first percentile of the pair (e.g., the 10^th percentile), upper whisker shows the latter percentile (e.g., the 90^th percentile).

   • 02-09th – 91-98th – The lower whisker will show the second percentile, the upper one the 98^th percentile. Apart from these two percentiles, the chart will also show two extra percentiles (the 9^th and 91^st) as two dots.

   • Outlier 1.5xIQR – IQR is a range between the 25^th and 75^th percentiles. The lower whisker shows the lowest datum still within 1.5 IQR of the lower quartile (25^th percentile - 1.5xIQR = lower inner fence). The upper whisker shows the highest datum still within 1.5 IQR of the upper quartile (75^th percentile + 1.5xIQR = upper inner fence). All values below lower OUTER fence or above upper OUTER fence are treated as, so called, outliers and plotted as filled dots. All values below lower INNER fence or above upper INNER fence are treated as, so called, suspected outliers and plotted as dots.

   • Outlier 3xIQR – IQR is a range between the 25^th and 75^th percentiles. The lower whisker shows the lowest datum still within 3 IQR of the lower quartile (25^th percentile - 3xIQR = lower outer fence). The upper whisker shows the highest datum still within 3 IQR of the upper quartile (75^th percentile + 3xIQR = upper outer fence). All values below lower outer fence or above upper outer fence are treated as so called outliers and plotted as filled dots.

10. Define the Sort By according to the axis X (alphabetically) or the axis Y (by measurement). Click the Ascending  icon to select whether the values will be sorted in ascending or descending order.

11. (Optional) You can define additional Box plot:

    1. Click the Add Measure to add a new Box plot.

    2. Select a Measure and an Expression.

    3. You can optionally show the measured value on the axis Z by checking the Z checkbox. This is necessary when the scale of the data of the different measurements on the Y-axis differ (e.g., cash value and percentage). Enabling this option here also automatically enables Show Z-axis on the Options tab.

    4. Repeat the process until you have defined all additional readings (bars, lines or splines).

12. (Optional) Click Add Series to be able to compare this chart with other data series. The selection of Data Source, all filters and the settings from the Data Display Options will be copied to the new series from Series 1.
    The data from all series are displayed in one chart and one data table. The maximum number of series in a chart is 10.

13. (Optional) In a configuration with multiple data series, you can compare series by creating a Join series. Set the relation (joining key) between the added data series and the base series. Note that series are always joined to the base Series 1, so you cannot compare, for example, Series 2 and 3. You can define up to 4 Join series. See Accelerate Actionable Insights for configuration examples.

14. Click the Apply and Refresh button or the  icon on the toolbar to update the chart.