Create, rename, duplicate and delete Views

When to use several Views on one dataset

The same dataset rarely serves one audience. A sales dataset might need a detailed analyst pipeline, a KPI summary for operations, a feed for a dashboard and a compliance export for finance. Without several Views, teams either copy data into separate files (version drift) or share one pipeline (conflicts).

Several Views give you:

• Parallel team work: two analysts build different pipelines on the same dataset at once. Changes in one View never affect another.
• Audience-specific outputs: each group gets data shaped for its needs, without restructuring or copying the dataset.
• Safe iteration: duplicate a working View to try new logic. If the experiment fails, delete it. The original is untouched.

Common View patterns

Name Views after their purpose, not their number, so your team knows what each one does.

Create a new View on a dataset

1. Open a dataset from the Data Library.
2. Click the + icon in the View tab bar at the bottom of the data editor.
3. The new View opens, starting with the full combined dataset and an empty pipeline.
4. Name it before you build the pipeline (see below).

You can also click New view in the dataset panel's icon strip. See The dataset panel.

Switch between Views in the data editor

The View tab bar runs along the bottom of the data editor. Each tab shows the View's name and, once the View has tasks, a pill with its task count (for example "3 Tasks"). Click a tab to open that View.

• Dataset settings: the icon at the start of the tab bar (tooltip "Manage dataset settings") opens the dataset's settings. See Dataset settings and history.
• All views (5): when the dataset has more than three Views, a list icon appears. It lists every View, with a check mark on the one that is open. Pick a View to jump to it.
• Reorder tabs: drag a tab left or right. The order is saved for you on this dataset. Other people keep their own order.
• View Settings: the More options (three-dot) button on a tab opens the View's menu. See Every option in the View Settings menu.

Name Views so your team can find them

View names are your team's main way to find a View. Use the pattern Audience or Purpose, then Destination or Frequency.

Prefix test and staging Views the same way every time, with TEST or STAGING, so people know not to rely on them for production outputs. Agree on the convention early: once a dataset has ten Views with mixed names, working out what each does means reading every pipeline.

Rename a View

Open the View's three-dot menu and choose Rename. The Rename view dialog opens. Rename is disabled while the pipeline has an error.

Duplicate a View to test changes safely

Duplicating creates an exact copy of a View: same pipeline tasks, configuration and settings. The copy is independent, so changes to it never affect the original.

Duplicate a View to:

• test new pipeline logic without touching a production View;
• build a manager summary starting from an analyst View;
• build a variant formatted for a different destination;
• isolate a time period by duplicating and changing the date filters.

1. Open the More options button on the View tab, or the three-dot menu on the View row in the dataset overview.
2. Choose Duplicate.

The copy appears next to the original, and a "View duplicated" message confirms it. Duplicate is disabled while the View's pipeline has an error, and in a View-only workspace. Fix the failing task first, then duplicate.

Never change a production View to test new logic. Duplicate it, prefix the copy with TEST, experiment, then promote the result or discard it.

Delete a View

1. Open the View's three-dot menu.
2. Choose Delete.
3. Confirm. The View moves to Trash, where it is deleted for good after 30 days unless you restore it.

Delete is disabled when the View is the dataset's only View: a dataset must always have at least one View. Before deleting, check that no team member or downstream destination relies on the View. For a View that is finished but may be needed later, rename it with an ARCHIVE prefix instead.

For what the delete dialog lists and what can block a delete, see Get things back from Trash.

Reset a View to start again

Reset removes everything applied to the View and returns it to a clean state. That includes all pipeline tasks, explore card filters, sorting in the data grid, column formats, display changes and anything else configured on the View.

1. Open the three-dot menu on the View row in the dataset overview (or on the View tab).
2. Choose Reset.
3. The Reset View dialog says "Resetting the View will remove all Transformations, Elements and Filters". Confirm.

This can't be undone. Use it when the pipeline no longer fits after a source schema change, or when you repurpose a View. To remove only display changes, such as column renames and formatting, and keep the pipeline, use the Display Changes panel at the bottom of the data editor instead. See How the pipeline works.

Save a View's pipeline as a template and reuse it

A template captures a View's pipeline as a set of rules, without the data. You can apply it to any View to repeat the same pipeline, on the same columns or different ones.

Save as a template: open the View's three-dot menu and choose Save as a template. The template downloads to your computer as a .mammoth file. It holds the full pipeline and can be shared, stored and reused across datasets and projects.

Apply template: open the View's three-dot menu and choose Apply template, then pick a template. The saved rules load into the current View. Apply template is active only when the View's pipeline is empty.

If the target dataset's columns differ, the template still applies, but tasks that reference missing columns show errors, for example "This column is missing: Department." Each errored task offers Fix or Delete. Work through them to fix mismatches or remove tasks that don't apply.

Use templates to:

• apply the same logic to datasets from similar source systems;
• set up a new dataset that has the same structure as an existing one;
• share a proven pipeline with other team members or projects as a file.

Save templates from proven Views, not works in progress. A template carries any errors in its pipeline to every View you apply it to.

Every option in the View Settings menu

Open the menu with the three-dot button on a View tab, or on a View row in the dataset overview. The menu is headed View Settings. Its items, top to bottom:

For what a scheduled email sends and how to set its timing, see Send messages. For the Auto-sync item, see Control when Views update with Auto-sync.

See who else is working in a View

When other people have the same View open, the bottom bar of the data editor shows their avatars. Hover the avatars to see their names, in a tooltip that ends "are working in this View". When there are too many names to list, it ends "and others are working in this View".

The avatars appear only when at least one other person has the View open. The list refreshes about once a minute. Agree with them before you change a pipeline someone else is working on, or use Draft mode so your changes stay pending until you apply them.

Keep each View to one purpose

A View that serves three audiences with a 40-task pipeline is harder to maintain, explain and change safely than three focused Views. When you can't describe a View in one sentence, split it.

Troubleshooting

Related

• The dataset panel: open a View and see its banners.
• How the pipeline works: the tasks inside a View, Draft mode and Display Changes.
• Pause sends while you edit: hold every destination at once.
• Get things back from Trash: restore a deleted View.
• Control when Views update with Auto-sync: Auto-sync, status chips and pending data.
• Dataset settings and history: the update method, separate from Auto-sync.