ever predicate: bool, [frame(...)]?

Keeps every row whose primary-key group had at least one row where the predicate evaluated to true inside the chosen time window.

On a Resource dataset, that keeps every interval for resources that matched the predicate at least once in scope; on other datasets with a primary key, a group is all rows sharing that key. The input must have a non-empty primary key (otherwise the compiler reports that the verb filters grouped rows and suggests filter instead). The predicate must be bool.

On temporal inputs without a second argument, the predicate is evaluated over the current query window. An optional frame limits how far backward and/or forward each row looks around its time; only back and/or ahead are allowed—start and end are rejected. For the opposite existential tests, use never or always.

Categories

• Filter

Accelerable

ever is accelerable if there is a frame() argument. A dataset that only uses accelerable verbs can be accelerated, making queries on the dataset respond faster.

Examples

ever string(status_code) ~ /^5.*/

Keeps every row in primary-key groups where the string form of status_code matched a server-error regex at least once over the query window, using ever without a frame() argument.

ever string(status_code) ~ /^5.*/, frame(back:30m)

Restricts the existential check to the trailing thirty minutes relative to each row’s time by supplying frame(back:30m) as the second argument.

ever bytesUsed > 20000000

Returns every row for primary-key groups where the numeric predicate was true on at least one row, including rows below the threshold, not only the rows that matched the predicate.