Altair® Panopticon

 

Aggregation Methods

Panopticon supports a wide range of aggregation methods. These methods are mathematical computations applied to a set of values. Values may include a group of numbers or numeric field values and variables. The following aggregation methods are available for most variables:

Aggregation Method

Description

Abs

The sum of absolutes of the selection.

Abs Sum

The absolute of the sum of the selection.

Combinations

Returns how many distinct combinations of breakdown column values there are below each node in the hierarchy

Count

The count of the number of rows in the selection.

Count Distinct

Creates numeric aggregated variables based on the distinct count of Text columns.

Count Non Zero

The count of non-zero values.

Cumulative Sum

The cumulative sum based on the currently applied sort order for each leaf node.

Cumulative Sum By Max

The cumulative sum of the sum of the value across siblings ordered by the max of the weight column.

Cumulative Sum Percent

Calculates the cumulative sum of items in a group, sorted by a specified numeric column, divided by the total sum of all items in the group. This result is the cumulative sum expressed as a decimal value between 0 and 1, which can be formatted and presented as a percentage value.

Do Not Aggregate

Returns the value of a single row, otherwise null.

External

Allows aggregates to be supplied from source data. The external aggregate configuration can be supplied explicitly, defined by the user, or implicitly from the data plug-in.

Harmonic Mean

The harmonic mean of the selection.

Intercept

The intercept of the least-squares line.

Level

The level in the hierarchy for the node or numbered from the leaf.

Max

The maximum value from the selection.

Mean

The mean of the selection.

Min

The minimum value from the selection.

Neg

The sum of the negative values in the selection.

Percentile

The selected percentile.

Percent of Parent

For each member item (child node) of a breakdown group (parent node), the percentage share of its value in relation to the parent group value, where the parent group value is calculated as the sum of all group member (child node) values:

[single child node value] / [sum of all child node values in the group]

The aggregate value is calculated as a ratio between 0 and 1 and will be presented as a percentage value by applying a percent format string in the aggregation settings.

Count Distinct Percent of Parent

Count Distinct expressed as a percentage share of the Count Distinct at the parent node level.

Percent Of Total

For each group and for each group member at all levels of the breakdown hierarchy, the percentage share of its value in relation to the total data set value, where the total is calculated as the sum across all rows in the dataset. This aggregate is similar to Percent of Parent, with the difference that the denominator or reference is ALWAYS based on the complete dataset:

[single node value] / [sum of all rows in the dataset]

The aggregate value is calculated as a ratio between 0 and 1 and will be presented as a percentage value by applying a percent format string in the aggregation settings.

Count Distinct Percent of Total

Count Distinct expressed as a percentage share of the Count Distinct in the total data table.

Percent of Total Change

This aggregate should be understood as “Change in (Percent of Total)”, not as “Percent of (Total Change)”. It is the result of calculating Percent of Total on two different columns, and then calculating the difference between them. The result is presented as the difference in percentage units, n.b.

This aggregate is typically used for comparing Percent of Total based on current values, to Percent of Total based on previous values. Therefore, the column specified as “Previous Values Column” in the settings, should be the column containing previous values.

The aggregate value is calculated as a ratio between 0 and 1 and will be presented as a percentage value by applying a percent format string in the aggregation settings. Optionally, you can emphasize that the value is a percentage units by customizing the format unit, for example: 0.00%'-units'.

Percent of Parent Reference

This aggregate works like Percent of Parent, with the difference that a value from one column is compared to a parent level sum of values from another column, which is set as the “Reference column”:

[single child node value from a column] / [sum of all child node values from the reference column in the group]

While Percent of Parent will always summarize to 100% at the group (parent) level, this is not the case with Percent of Parent Reference, which can summarize to any number, depending on the differences between the Values and the Reference Values.

The aggregate value is calculated as a ratio between 0 and 1 and will be presented as a percentage value by applying a percent format string in the aggregation settings.

Percent Of Total Reference

This aggregate works like Percent of Total, with the difference that a value from one column is compared to a total data set level sum of values from another column, which is set as the “Weight column”:

[single node value from a column] / [sum of all rows from the reference column in the dataset]

While Percent of Total will always summarize to 100% across the whole data set, this is not the case with Percent of Total Reference, which can summarize to any number, depending on the differences between the Values and the Reference Values.

The aggregate value is calculated as a ratio between 0 and 1 and will be presented as a percentage value by applying a percent format string in the aggregation settings.

Pos

The sum of the positive values in the selection

Product

The product of the selection.

Ratio of Sums

The comparison between the sum of a selected measure divided by the sum of the selected reference measure.

Sibling Rank

The numeric rank of siblings within a hierarchy branch.

Sibling Rank Percent

The numeric rank of each sibling (items in the same group or category), divided by the total count of siblings. Possible values include decimal values > 0 and <= 1.

Slope

The slope of the least-squares line.

Stdev

The Standard Deviation of the selection.

Stdevp

The Population Standard Deviation of the selection.

Sum

The sum of the selection.

Unique

Used with numeric values and will display a number in case all the values in a group are the same , otherwise it will show empty/null. This aggregation can be used as an indicator of a logical test: “if the numeric values in this group and in any subgroups are identical, then show the numeric value, or else show nothing”.

Text Concat Distinct

Aggregates text fields to display all possible text values in a comma delimited list.

Text Unique

Aggregates text fields to display distinct values.

Weighted Mean

The weighted mean of the selection, based on a specified weighting column.

Weighted Harmonic Mean

The weighted harmonic mean of the selection, based on a specified weighting column.

Weighted Sum

The sum of the product of the selected field and the weight field.