Dimensions and KPIs
About dimensions and KPIs in analysis components
Analyses components require the configuration of Dimensions and Key Performance Indicators (KPIs).
Dimensions basically represent the columns of the table you selected as data source for the analyses.
KPIs are functions that consolidate a set of values belonging to a single occurrence inside a dimension into one single value.
Dimensions and KPIs can be set by the Analyst and worked with by Viewer and Analyst.
Detailed Information
Dimension
Dimensions basically represent the columns of the table you selected as the data source for this document. In contrast to KPIs, there are no restrictions as to which data types can serve as dimensions therefore all columns available in the table can be selected. Your choice of a dimension will define the possible level of KPIs for your data. If, for example, you add a dimension containing dates and want to round all dates to month, there will be a maximum of twelve possible KPIs for one year. However, if you round the dimension to years, there will – for one year – only be one entry with one KPI in your analysis component.
Numbers
There is no limit to dimensions, however, you should keep in mind that too many dimensions might increase the complexity of the component for a viewer of your analysis.
Example
If, for example, you add a dimension containing dates and want to round all dates to month, there will be a maximum of twelve possible KPIs for one year. However, if you round the dimension to years, there will – for one year – only be one entry with one KPI in your analysis component.
KPIs
KPIs are functions that consolidate a set of values belonging to a single occurrence inside a dimension into one single value. Consolidation can be done by accumulating the values, by calculating the average, minimum, or maximum, or simply by counting the number of occurrences.
Example
To give a short example, let's assume your data contains a table listing all invoices you received from vendors and their respective order values. Now if you choose "vendor" as a dimension and the sum of "order value" as KPI, for each vendor all entries in your data will be accumulated regarding their order values. Your result set will contain one entry for each vendor and their respective sum of order values. If you choose "avg" as an aggregation function, the average of the order values will be calculated for each vendor. If you choose "min" or "max", the minimum or maximum order value will be selected and presented in the result set.
Apart from the function "count", all KPI functions need to be based on another column than the dimension column. Since these functions can only be performed on numerical data, only columns containing numerical data types will be available for selection. Basically, the table selected as a data source will be scanned for numerical data types and all columns meeting the criteria will be provided as bases for KPIs. The function "count" simply counts the number of occurrences for each value in the dimensions column (so for our vendor example, the result set would contain an entry for each vendor and the number of invoices you received from them). If you use KPIs with two or more dimensions KPIs will be calculated for each unique combination of all dimension values occurring in the data source.
Chart Types
Depending on the Chart Type, the number of possible dimensions and KPIs will vary.
Please see the following table to have an overview of possibilities how to apply Dimensions and KPIs
Chart Type | Dimensions | KPIs |
|---|---|---|
OLAP Table | ∞ | ∞ |
Pivot Table | ∞ | ∞ |
Column Chart | 1 | ∞ |
Pie Chart | 1 | |
Donut Chart | 1 | |
Line Chart | 1 | ∞ |
Area Chart | 1 | ∞ |
Scatter Plot | 1 or 2 | 1 or 0 |
Bubble Plot | 1 | 2 |
Boxplot | 1 | 1 |
Worlmap | 1 | 1 |