Skip to content
DataSquaresDocs

Fixed

Context

Fixed(dim…, expr [, IgnoreFilters])

Evaluates expr grouped by exactly the listed dimensions, then joins the result back to the chart’s grain. Dimensions come first, the expression last (SQX007). Date columns take an inline grain: Fixed(Date.Date:month, …).

Filters just work: unlike Tableau’s FIXED, SquareX Fixed evaluates inside the active filter context — no context-filter promotion needed. When you genuinely want the filter-independent number, add the explicit IgnoreFilters flag as the last argument: in Fixed(Customer.Id, [Revenue], IgnoreFilters) each customer’s value is computed ignoring every user filter, while which customers appear still follows the view (and row-level security always applies).

When the chart is coarser than the Fixed grain, the chart’s field-well aggregation re-aggregates the joined-back values — drop [CustomerRevenue] on a by-region chart with agg avg and you get average revenue per customer per region.

Composes with Calculate, both ways. Wrapping (Calculate(Fixed(…), mods)) evaluates the whole fixed computation — which dim values exist AND their values — under the modified context; removing a dimension makes the fixed value replicate across it, and a DateShift on the wrapper gives prior-period fixed grains. Putting the Calculate (or a measure that is one) inside the body (Fixed(Customer.Id, [WestRevenue])) modifies only the per-dim values — the set of customers still comes from the unmodified context.

When to use it

Section titled “When to use it”
  • Per-entity values reused at any grouping: revenue per customer, units per store, first/latest order date per account.
  • Distinct counts at a pinned grain: Fixed(Date.Date:month, CountDistinct(...)) for customers-per-month.
  • Denominators pinned to a grain regardless of what the viewer groups by — if the grain should follow the view instead, use Include/Exclude.

Example

Section titled “Example”
CustomerRevenue := Fixed(Customer.Id, [Revenue])


CustomersPerMonth := Fixed(Date.Date:month, CountDistinct(Sales.CustomerId))


// West-only per-customer revenue — chart with avg for "avg revenue
// per customer, West only" at any grouping
WestRevPerCustomer := Calculate(Fixed(Customer.Id, [Revenue]), Sales.Region = "West")

Under the hood

Section titled “Under the hood”

The fixed grain computes in its own CTE, then joins back to the chart’s rows:

// On a bar chart grouped by region, field-well aggregation = avg
CustomerRevenue := Fixed(Customer.Id, [Revenue])

At a different grain than the chart, the engine inserts a bridge CTE so each customer counts once per group — deliberately not Tableau’s row-replication.

Good to know

Section titled “Good to know”
  • Join-backs are LEFT JOINs: outer rows with no matching fixed group carry blank and outer re-aggregation skips them.
  • In-language re-aggregation (Avg([CustomerRevenue])) is not expressible yet — aggregation arguments are columns (SQX013); set the aggregation in the chart instead.
  • A DateShift belongs on a Calculate wrapping the Fixed — inside the body it would displace the dim domain and the values inconsistently (SQX010). Nested Fixed and YTD/RollingSum inside a Fixed body are also SQX010.
Section titled “Related”
© 2026 DataSquares