AVERAGE

In the Ultra Transcenders books

PL-300

Each book explains AVERAGE in context, with comparison tables and the common traps.

Terms in this definition

• DAX

  Short for Data Analysis Expressions, the formula language you write measures, calculated columns and row-level security rules in, within tabular semantic models in Power BI, Excel Power Pivot or Analysis Services.

• BLANK

  DAX uses BLANK to mean no value at all. It resembles NULL in SQL but behaves more like an empty cell in Excel, so adding 20 to it returns 20; the BLANK function creates one and ISBLANK detects it.

• AVERAGEX

  Returns the mean of an expression after computing it once for each row in a table; if the table has no rows, this DAX iterator gives BLANK.

Related terms

• Automatic statistics

  When queries run against the Fabric Data Warehouse or the SQL analytics endpoint, the engine creates statistics (histograms, cardinality and average column length) on columns used for filtering, joining, grouping, sorting or DISTINCT. The histogram objects are named with a _WA_Sys_ prefix.

• Custom totals

  Lets a table or matrix column total show something other than the default, such as an average, minimum, maximum, count, distinct count or no total, based on the rows shown; you pick Customize total calculation. Power BI builds this with a visual calculation and appends _Base to the source column's name.

• Default summarization

  A setting on a numeric column that decides which aggregation, such as Sum or Average, Power BI applies when the column is dropped into a visual. Identifier and year columns are best set to Don't summarize.

• GitHub Actions metrics

  Found in organisation Insights. Usage figures report minutes used by workflows and jobs; performance figures report average run and queue times and how often jobs fail.

• Implicit measure

  A summary, such as Sum, Average or Count, that Power BI builds on its own when a column is dropped into a visual. Excel PivotTables and other MDX clients, as well as calculation groups, rely on explicit measures instead, and the Discourage implicit measures setting prevents them.

• L1 test

  Like L0, it depends only on the assembly under test, but can be slower: an average below 400 ms, none above 2 s.

• MTTR

  Mean time to recovery: how long, on average, it takes to get a service working again after an outage. It is among the DORA delivery measures, where it is called time to restore service.

• RMSE

  Root mean squared error, a regression metric: take the differences between predicted and actual values, square them, average them and take the square root. Smaller values are better.