Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.siftstack.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

This guide walks you through the process of creating a Sift Family. Use Families to group a set of related Runs so you can compare them side-by-side, compute aggregate statistics, and build rules that flag anomalies against your historical baseline. For example, all nominal tests for a given engine configuration, or daily Runs from a deployed vehicle.

​
Prerequisites

  • You have at least two Runs ingested into Sift that you want to group.
  • Your Sift user account belongs to a Sift group with the Editor role on the organization containing the Runs.

​
Create a Family

To create a Family:
  1. In the Sift navigation panel, click Families. On smaller screens, Families may not be immediately visible — scroll right within the navigation panel to find it.
  2. Click Create family.
  3. Follow the Create Family steps, and then click Save.

​
Basic Info

Enter the following details for the Family:

​
Membership

Add the Runs for this Family. Excluding a Run keeps it in the Family but removes it from analysis and visualization. Use exclusion for known outliers or failed tests that you want to track alongside the Family but not include in aggregates. You can always add or remove members later from the Family management page.

​
Candidate Runs

Use a candidate query to find Runs that match criteria instead of selecting them one at a time. A candidate query is a CEL expression that is evaluated against each Run. Matching Runs appear in the Candidates tab, where you can bulk include them in the Family or exclude them.
  1. In the Members step, click the Candidates tab.
  2. Click Add candidate query.
  3. Enter a CEL expression. For example, to match Runs whose name contains flight for a specific Asset: 
    name.contains('flight') && asset_name == 'rover-1'
  4. Click Run Query to see matching Runs.
  5. Select Runs and apply the Include or Exclude bulk action. Included Runs move to the Family Members tab; excluded Runs move to the Excluded tab.
Click Show CEL documentation in the editor for the full list of available fields, operators, and example expressions. To reuse a query from another Family, click Import query from existing Family and select the source Family. The query opens in edit mode and you can adjust the query before running it. The candidate query is saved with the Family. Any Run—newly ingested or existing—that matches the query, and is not already included in or excluded from the Family, appears as a review candidate on the Family overview page. Well-scoped queries, combined with explicit inclusions and exclusions, let the candidate list signal that a new Run is a valid member for the Family. See Review candidate Runs for details.

​
Align Data

Alignments define how Runs are time-synchronized during comparison. Without an alignment, each Run’s time axis starts at its own absolute start time, making cross-Run comparison difficult. Configure at least one alignment: You can define multiple alignments and switch between them in Explore. Alignment names must be unique within the Family. Alignments can be edited after the Family is saved.

​
Statistics (optional)

Family statistics let you compute aggregate time series across all included members — for example, the mean and standard deviation of a channel across all Runs. You can configure statistics now or add them later from the Family management page. See Configure Family statistics for details.