> ## Documentation Index
> Fetch the complete documentation index at: https://honeydew.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Sigma

Honeydew can publish a domain to Sigma as a data model, so you can build workbooks on
top of the Honeydew semantic layer. The data model is backed by a Honeydew connection
in Sigma.

<Info>
  Sigma support is in Beta.
</Info>

<Note>
  Publishing a data model requires a one-time API access setup — see
  [Sigma API Integration](/docs/integration/bi-tools/sigma-api-integration).
</Note>

## Publishing a domain as a data model

From the Honeydew UI, select **Publish data model** and provide:

1. **Domain** — the domain to publish as the data model.
2. **Workspace** — the Sigma workspace to place the data model in.
3. **Connection** — the Honeydew connection in Sigma (only selectable when more than one exists).
4. **Folder** — the folder for the data model.
5. **Data model name** — type a name to create a new data model, or select an existing
   one in the folder to update. See [Updating an existing data model](#updating-an-existing-data-model).
6. **Version tags** — optional Sigma version tags to apply to the published version.
   See [Version tags](#version-tags).

Honeydew creates (or updates) the Sigma data model backed by the Honeydew connection.
Along with the fields, Honeydew syncs the following metadata:

* Field type — every attribute is synced as a dimension, and every metric as a measure.
* Folders — dimensions are grouped into folders in Sigma.
* Field display names. When no display name is set, Honeydew creates an automatic
  user friendly name.
* Field descriptions.
* Field format string (see [Sigma-specific Metadata](#sigma-specific-metadata)).

<Tip>
  You can now build workbooks on top of the published data model, using the Sigma user interface.
</Tip>

### Updating an existing data model

The **Data model name** field is where you choose between creating and updating:

* Typing a new name creates a new data model (marked **New**).
* Selecting an existing model in the folder updates it (marked **Override**). Honeydew
  shows the version tags currently applied to that model, and asks you to confirm before
  overriding it.

Each publish adds a new version to the data model, keeping earlier versions in Sigma's
version history.

### Version tags

Version tags let you pin a published version so workbook builders can consume a specific
version instead of the latest. Select one or more of your organization's
[Sigma version tags](https://help.sigmacomputing.com/docs/version-tagging) to apply to
the version being published.

A version tag moves to the latest published version each time you apply it, so a tag such
as `Production` always points to the most recently blessed version.

<Note>
  Tags must already exist in Sigma before you can apply them. Applying a tag is best-effort:
  if the model publishes but tagging fails, Honeydew reports that tagging failed and still
  keeps the published version.
</Note>

### Sigma-specific Metadata

You can control how a field is formatted in Sigma by adding a `formatString` to the
field's `sigma` metadata section. The value is a D3 format string, which Sigma applies
natively (for example `$,.2f` for currency or `,.0f` for a whole number). See Sigma's
[format columns and metrics](https://help.sigmacomputing.com/docs/format-columns-and-metrics-in-the-code-representation-of-a-data-model)
documentation.

<Note>
  Only number and date/time fields can be formatted. A `formatString` on a string or
  boolean field is ignored.
</Note>

#### Metrics and Calculated Attributes

For [metrics](/docs/calculations/metrics#yaml-schema) and
[calculated attributes](/docs/calculations/attributes#yaml-schema),
add the metadata to the field's own YAML file:

```yaml theme={null}
...
metadata:
  - name: sigma
    metadata:
      - name: formatString
        value: '$,.2f'
```

For [dataset attributes](/docs/modeling/source-data#yaml-schema), add the metadata under
the attribute entry in the dataset's YAML file:

```yaml theme={null}
...
attributes:
  ...
  - column: <src column name>
    name: <attribute name>
    ...
    metadata:
      - name: sigma
        metadata:
          - name: formatString
            value: ',.0f'
```

## Refreshing a data model

Changes in calculations reflect automatically and immediately in Sigma.

<Tip>
  Connect Sigma to a development branch so logic changes in Honeydew reflect
  immediately in your workbook — a quick experimentation workflow on Honeydew semantics.
</Tip>

Any other change will not reflect in Sigma until publishing the data model again.

In particular, the following changes require publishing again:

1. Changing metadata: descriptions and display names
2. Adding new attributes or metrics
3. Renaming or removing attributes or metrics
4. Changing the entities or fields that participate in a domain