Skip to main content
Version: Next

Build Your First Subject

Goal​

Create your first Subject for analysis: pick topics, define columns, filters and joins, then explore the dataset and create reports.

Prerequisites​

  • Topics available in a connected space.
  • Console access to create and view subjects.

1) Pick Topics​

  • Open Subject and start the wizard.
  • Step 1: pick one or more topics from the connected space to form the subject’s data scope.
  • Arrange selections so that joins make sense if multiple topics are chosen.

2) Define Columns​

  • Step 2: add columns for the subject.
  • A column can be:
    • From a topic factor directly,
    • A constant value,
    • Computed via functions (math/date/case‑when, etc.).
  • Give friendly names to columns for readability.

3) Define Filters​

  • Step 3: add filters applied to all reports built on this subject.
  • Use filters to narrow the dataset to the relevant slice (time window, status, id ranges, etc.).

4) Define Joins​

  • Step 4: when two or more topics are included, define joins between them.
  • Join types:
    • Exactly Match: intersection of both sides,
    • Left Side Prioritized: left side plus intersection,
    • Right Side Prioritized: right side plus intersection.
  • Skip joins when only one topic is used.

5) View Definition DSL​

  • Review the YAML‑like DSL of the subject definition.
  • Resolve any issues highlighted in the wizard.

6) Dataset​

  • Open the dataset page to preview data.
  • Use pagination and the top‑right page switcher.
  • Columns can be resized, sorted, drag‑and‑dropped and locked to the left.
  • Validate that the dataset matches expectations before building reports.

7) Reports​

  • Switch to the reports page.
  • Create new reports based on the subject: choose charts, configure fields and filters.
  • Iterate until the visualizations reflect your analysis goals.

Header Actions​

  • Back to connected space.
  • Switch between structure, dataset and reports.
  • Create new report, switch subject or delete current one.
  • Rename the subject by clicking its name in the header.

Tips​

  • Keep column names concise and meaningful.
  • Start with one topic, verify results, then add joins for multi‑topic analysis.
  • Apply global filters at the subject level to ensure consistent reporting.
  • Use computed columns for derived metrics and date bucketing.

What’s Next​

  • Add more subjects for different perspectives.
  • Share common filters and computed columns across reports.
  • Export datasets for downstream analytics or AI workflows.