The platform's charts automatically calculate averages and display those averages in the charts to facilitate user understanding and analysis. The platform calculates average lines differently depending on Proof's chart-metric classification system. This helps ensure logical and useful averages are displayed based on the chart and metric type.
Classification System
The average calculations are determined based on a three-level classification system.
1. Groupings
There are three primary groupings of chart types that help to organize the different charts in the platform:
Value charts that show a single value.
By Data Submitter charts that show each data submitting entity on the x-axis.
By Reporting Period charts that show each reporting period on the x-axis.
These groupings are typically identified in the name of the chart (e.g., "Bar - by Data Submitter").
2. Visualization Categories
Within each of these broad groupings of chart types, there are three different visualization categories that also influence how the average line is calculated:
Absolute data that show information on performance that is usually collected in terms of absolute units of measurement over a given period of time (e.g., Scope 1 emissions as measured in metric tons).
Normalized data create relationships between figures visible, by relating two absolute figures to each other (e.g., Scope 1 emission intensity, calculated by dividing Scope 1 emissions over total revenue).
Point In Time data have distinct or unique units of measure that should not be summed over time (e.g., total number of employees, where employees may be the same from one year to the next).
Visualization categories are determined in Proof's backend data infrastructure.
3. Metric Types
Finally, there are two metric types that may exist within each visualization category:
Base metrics are singular metrics that are not made up of any other metrics (e.g., Scope 1 emissions).
Composed metrics are calculated from multiple base metrics (e.g., Scope 1 emission intensity, which is derived by dividing Scope 1 emissions by total revenue).
Metric types are determined in Proof's backend data infrastructure.
Average Calculations, by Classification Level
The table below displays the breakdown of how average lines are calculated across each variation of the three classifications above:
Grouping | Visualization Category | Metric Type | Average Calculation |
Value | Normalized | Base | Simple average |
Value | Normalized | Composed | Weighted average* |
By Data Submitter | Absolute | Base | Simple average |
By Data Submitter | Absolute | Composed | Weighted average/# of data submitters** |
By Data Submitter | Normalized | Composed | Weighted average |
By Data Submitter | Point In Time | Base | Simple average |
By Data Submitter | Point In Time | Composed | Simple average |
By Reporting Period | Absolute | Base | Simple average/# of reporting periods |
By Reporting Period | Absolute | Composed | Weighted average/# of reporting periods** |
By Reporting Period | Normalized | Composed | Weighted average |
By Reporting Period | Point In Time | Base | Simple average/# of reporting periods |
By Reporting Period | Point In Time | Composed | Weighted average/# of reporting periods** |
*Weighted averages are calculated as the sum of numerators over sum of denominators, for data where the numerator and denominator are both present. This helps reduce bias from small sample sizes when calculating averages within groups.
**This average calculation may vary by metric depending on the applicable formula for the composed metric.
For further information or support in determining how a specific average line is calculated, contact us at [email protected].