SHOW COLUMNS IN SEMANTIC VIEW¶

Lists all queryable columns in a semantic view – dimensions, facts, and metrics – with their data types, expressions, kind, and comments. Private items are excluded from the output.

Syntax¶

SHOW COLUMNS IN SEMANTIC VIEW <name>

Parameters¶

<name>: The name of the semantic view. Returns an error if the view does not exist.

Output Columns¶

Returns one row per queryable column with 8 columns:

Column	Type	Description
`database_name`	VARCHAR	The DuckDB database containing the semantic view.
`schema_name`	VARCHAR	The DuckDB schema containing the semantic view.
`semantic_view_name`	VARCHAR	The semantic view name.
`column_name`	VARCHAR	The dimension, fact, or metric name.
`data_type`	VARCHAR	The declared output type. Empty for every view created since v0.10.0 – no surface can declare a type and nothing infers one. Populated only for views stored before that release. See Reported Data Types.
`kind`	VARCHAR	The column kind: `DIMENSION`, `FACT`, `METRIC`, or `DERIVED_METRIC`.
`expression`	VARCHAR	The SQL expression defining the column.
`comment`	VARCHAR	The comment text. Empty string if no comment is set.

Kind Values¶

Kind	Description
`DIMENSION`	A grouping expression from the `DIMENSIONS` clause.
`FACT`	A row-level expression from the `FACTS` clause.
`METRIC`	A base metric (scoped to a table) from the `METRICS` clause.
`DERIVED_METRIC`	A derived metric (no table alias, references other metrics) from the `METRICS` clause.

PRIVATE Exclusion¶

Metrics and facts marked PRIVATE are excluded from the output. Only PUBLIC items (the default) appear. This matches the behavior of wildcard expansion in semantic_view() queries.

Sort Order¶

Rows are sorted by kind (alphabetically: DERIVED_METRIC, DIMENSION, FACT, METRIC) and then by column_name within each kind.

Examples¶

CREATE SEMANTIC VIEW shop AS
TABLES (o AS orders PRIMARY KEY (id))
FACTS (o.raw_amount AS o.quantity * o.price COMMENT = 'Line total')
DIMENSIONS (o.region AS o.region)
METRICS (
    o.revenue     AS SUM(o.quantity * o.price),
    o.order_count AS COUNT(o.id),
    avg_order     AS revenue / order_count
);

SHOW COLUMNS IN SEMANTIC VIEW shop;

┌───────────────┬─────────────┬────────────────────┬─────────────┬───────────┬────────────────┬───────────────────────────┬────────────┐
│ database_name │ schema_name │ semantic_view_name │ column_name │ data_type │ kind           │ expression                │ comment    │
├───────────────┼─────────────┼────────────────────┼─────────────┼───────────┼────────────────┼───────────────────────────┼────────────┤
│ memory        │ main        │ shop               │ avg_order   │           │ DERIVED_METRIC │ revenue / order_count     │            │
│ memory        │ main        │ shop               │ region      │           │ DIMENSION      │ o.region                  │            │
│ memory        │ main        │ shop               │ raw_amount  │           │ FACT           │ o.quantity * o.price      │ Line total │
│ memory        │ main        │ shop               │ order_count │           │ METRIC         │ COUNT(o.id)               │            │
│ memory        │ main        │ shop               │ revenue     │           │ METRIC         │ SUM(o.quantity * o.price) │            │
└───────────────┴─────────────┴────────────────────┴─────────────┴───────────┴────────────────┴───────────────────────────┴────────────┘

avg_order is a DERIVED_METRIC because it has no table alias and its expression names two other metrics. A derived metric may not contain an aggregate function of its own – avg_order AS revenue / COUNT(*) is rejected at CREATE, which is why the order count is declared as the base metric o.order_count first. See How to Compose Metrics with Derived Metrics.

data_type is empty for every row here because no surface can declare a member’s output type: the SQL DDL has no clause for it, and the YAML output_type field was withdrawn because GET_DDL could not carry it (a restored view silently lost the cast). Nothing infers one either – v0.10.0 removed the define-time inference pass – so the column is populated only for views stored before that release. See Reported Data Types.

Error: view does not exist:

SHOW COLUMNS IN SEMANTIC VIEW nonexistent;

Error: semantic view 'nonexistent' does not exist