Share via

CREATE VIEW

Applies to: check marked yes Databricks SQL check marked yes Databricks Runtime

Constructs a virtual table that has no physical data based on the result-set of a SQL query or a metric view based on a yaml specification. ALTER VIEW and DROP VIEW only change metadata.

To execute this statement, you must be a metastore administrator or have USE CATALOG and USE SCHEMA privileges on the catalog and schema, along with CREATE TABLE privileges in the target schema.

The user executing this command will become the owner of the view.

Syntax

CREATE [ OR REPLACE ] [ TEMPORARY ] VIEW [ IF NOT EXISTS ] view_name
    [ column_list ]
    [ with_clause |
      COMMENT view_comment |
      DEFAULT COLLATION collation_name |
      TBLPROPERTIES clause |
      LANGUAGE YAML ] [...]
    AS { query | $$ yaml_string $$ }

with_clause
   WITH { { schema_binding | METRICS } |
          ( { schema_binding | METRICS } [, ...] } )

schema_binding
   WITH SCHEMA { BINDING | COMPENSATION | [ TYPE ] EVOLUTION }

column_list
   ( { column_alias [ COMMENT column_comment ] } [, ...] )

Parameters

  • OR REPLACE

    If a view of the same name already exists, it is replaced. To replace an existing view you must be its owner.

    Replacing an existing view does not preserve privileges granted on the original view. Use ALTER VIEW to preserve privileges.

    CREATE OR REPLACE VIEW view_name is equivalent to DROP VIEW IF EXISTS view_name followed by CREATE VIEW view_name.

  • TEMPORARY

    TEMPORARY views are visible only to the session that created them and are dropped when the session ends.

  • GLOBAL TEMPORARY

    Applies to: check marked yes Databricks Runtime

    GLOBAL TEMPORARY views are tied to a system preserved temporary schema global_temp.

  • IF NOT EXISTS

    Creates the view only if it does not exist. If a view by this name already exists the CREATE VIEW statement is ignored.

    You may specify at most one of IF NOT EXISTS or OR REPLACE.

  • view_name

    The name of the newly created view. A temporary view's name must not be qualified. The fully qualified view name must be unique.

    View names created in hive_metastore can only contain alphanumeric ASCII characters and underscores (INVALID_SCHEMA_OR_RELATION_NAME).

  • METRICS

    Applies to: check marked yes Databricks SQL check marked yes Databricks Runtime 16.4 and above check marked yes Unity Catalog only

    Identifies the view as a metric view. The view must be defined with LANGUAGE YAML and the body of the view must be a valid yaml specification.

    This clause is not supported for temporary views.

    A metric view does not support the DEFAULT COLLATION and schema_binding clauses.

    The YAML specification of the metric view defines dimensions and measures. The dimensions are the columns of the view by which the invoker can aggregate the measures, while the measures define the aggregations of the view.

    The invoker of a metric view uses the measure expression to access the views defined measures instead of specifying aggregation functions.

  • schema_binding

    Applies to: check marked yes Databricks Runtime 15.3 and above

    Optionally specifies how the view adapts to changes to the schema of the query due to changes in the underlying object definitions.

    This clause is not supported for temporary views, metric views, or materialized views.

    • SCHEMA BINDING

      The view will become invalid if the query column-list changes except for the following conditions:

      • The column-list includes a star clause, and there are additional columns. These additional columns are ignored.
      • The type of one or more columns changed in a way that allows them to be safely cast to the original column types using implicit casting rules.

      This is the default behavior.

    • SCHEMA COMPENSATION

      The view will become invalid if the query column list changes except for the following conditions:

      • The column-list includes a star clause, and there are additional columns. These additional columns are ignored.
      • The type of one or more columns changed in a way that allows them to be cast to the original column types using explicit ANSI cast rules.

    • SCHEMA TYPE EVOLUTION

      The view will adopt any changes to types in the query column list into its own definition when the SQL compiler detects such a change in response to a reference to the view.

    • SCHEMA EVOLUTION

      • This mode behaves like SCHEMA TYPE EVOLUTION, and also adopts changes in column names or added and dropped columns if the view does not include an explicit column_list.
      • The view will only become invalid if the query can no longer be parsed, or the optional view column_list does not match the number of expressions in the query select-list anymore.

  • column_list

    Optionally labels the columns in the query result of the view. If you provide a column list the number of column aliases must match the number of expressions in the query or, for metric views, the YAML specification. In case no column list is specified aliases are derived from the body of the view.

    • column_alias

      The column aliases must be unique.

    • column_comment

      An optional STRING literal describing the column alias.

  • view_comment

    An optional STRING literal providing a view-level comments.

  • DEFAULT COLLATION collation_name

    Applies to: check marked yes Databricks SQL check marked yes Databricks Runtime 16.3 and above

    Defines the default collation to use within query. If not specified, the default collation is UTF8_BINARYa.

    This clause is not supported for metric views.

  • TBLPROPERTIES

    Optionally sets one or more user defined properties.

  • AS query

    A query that constructs the view from base tables or other views.

    This clause is not supported for metric views.

  • AS $$ yaml_string $$

    A yaml specification that defines a metric view.

Examples

-- Create or replace view for `experienced_employee` with comments.
> CREATE OR REPLACE VIEW experienced_employee
    (id COMMENT 'Unique identification number', Name)
    COMMENT 'View for experienced employees'
    AS SELECT id, name
         FROM all_employee
        WHERE working_years > 5;

-- Create a temporary view `subscribed_movies`.
> CREATE TEMPORARY VIEW subscribed_movies
    AS SELECT mo.member_id, mb.full_name, mo.movie_title
         FROM movies AS mo
         INNER JOIN members AS mb
            ON mo.member_id = mb.id;

-- Create a view with schema binding (default)
> CREATE TABLE emp(name STRING, income INT);
> CREATE VIEW emp_v WITH SCHEMA BINDING AS SELECT * FROM emp;

– The view ignores adding a column to the base table
> ALTER TABLE emp ADD COLUMN bonus SMALLINT;
> SELECT * FROM emp_v;
name  income
----  ------

-- The view tolerates narrowing the underlying type
> CREATE OR REPLACE TABLE emp(name STRING, income SMALLINT, bonus SMALLINT);
> SELECT typeof(income) FROM emp_v;
 INTEGER

– The view does not tolerate widening the underlying type
CREATE OR REPLACE TABLE emp(name STRING, income BIGINT, bonus SMALLINT);
> SELECT typeof(income) FROM emp_v;
 Error

– Create a view with SCHEMA COMPENSATION
> CREATE TABLE emp(name STRING, income SMALLINT, bonus SMALLINT);
> CREATE VIEW emp_v WITH SCHEMA COMPENSATION AS SELECT * FROM emp;

-- The view tolerates widening the underlying type but keeps its own signature fixed
CREATE OR REPLACE TABLE emp(name STRING, income INTEGER, bonus INTEGER);
> SELECT typeof(income) FROM emp_v;
 INTEGER

-- The view does not tolerate dropping a needed column
ALTER TABLE emp DROP COLUMN bonus;
> SELECT * FROM emp_v;
Error

– Create a view with SCHEMA EVOLUTION
> CREATE TABLE emp(name STRING, income SMALLINT);
> CREATE VIEW emp_v WITH SCHEMA EVOLUTION AS SELECT * FROM emp;

-- The view picks up additional columns
> ALTER TABLE emp ADD COLUMN bonus SMALLINT
> SELECT * FROM emp_v;
 name income bonus
 ---- ------ -----

-- The view picks up renamed columns as well
> ALTER TABLE emp RENAME COLUMN income TO salary SMALLINT;
> SELECT * FROM emp_v;
 name salary bonus
 ---- ------ -----

-- The view picks up changes to column types and dropped columns
> CREATE OR REPLACE TABLE emp(name STRING, salary BIGINT);
> SELECT *, typeof(salary)AS salary_type FROM emp_v;
 name salary
 ---- ------

-- Create a view using a default collation of UTF8_BINARY
> CREATE VIEW v DEFAULT COLLATION UTF8_BINARY
    AS SELECT 5::STRING AS text;

-- Creates a Metric View as specified in the YAML definition, with three dimensions and four measures representing the count of orders.
> CREATE OR REPLACE VIEW region_sales_metrics
  (month COMMENT 'Month order was made',
   status,
   order_priority,
   count_orders COMMENT 'Count of orders',
   total_Revenue,
   total_Revenue_p_Customer,
   total_revenue_for_open_orders)
  WITH METRICS
  LANGUAGE YAML
  COMMENT 'A Metric View for regional sales metrics.'
  AS $$
   version: 0.1
   source: samples.tpch.orders
   filter: o_orderdate > '1990-01-01'
   dimensions:
   - name: month
     expr: date_trunc('MONTH', o_orderdate)
   - name: status
     expr: case
       when o_orderstatus = 'O' then 'Open'
       when o_orderstatus = 'P' then 'Processing'
       when o_orderstatus = 'F' then 'Fulfilled'
       end
   - name: prder_priority
     expr: split(o_orderpriority, '-')[1]
   measures:
   - name: count_orders
     expr: count(1)
   - name: total_revenue
     expr: SUM(o_totalprice)
   - name: total_revenue_per_customer
     expr: SUM(o_totalprice) / count(distinct o_custkey)
   - name: total_revenue_for_open_orders
     expr: SUM(o_totalprice) filter (where o_orderstatus='O')
  $$;

> DESCRIBE EXTENDED region_sales_metrics;
  col_name                       data_type
  ------------------------------ --------------------------
  month                          timestamp
  status                         string
  order_priority                 string
  count_orders                   bigint measure
  total_revenue                  decimal(28,2) measure
  total_revenue_p_customer       decimal(38,12) measure
  total_revenue_for_open_orders  decimal(28,2) measure

  # Detailed Table Information
  Catalog                        main
  Database                       default
  Table                          region_sales_metrics
  Owner                          alf@melmak.et
  Created Time                   Thu May 15 13:03:01 UTC 2025
  Last Access                    UNKNOWN
  Created By                     Spark
  Type                           METRIC_VIEW
  Comment                        A Metric View for regional sales metrics.
  Use Remote Filtering           false
  View Text                      "
     version: 0.1
     source: samples.tpch.orders
     filter: o_orderdate > '1990-01-01'
     dimensions:
     - name: month
       expr: date_trunc('MONTH', o_orderdate)
     - name: status
       expr: case
         when o_orderstatus = 'O' then 'Open'
         when o_orderstatus = 'P' then 'Processing'
         when o_orderstatus = 'F' then 'Fulfilled'
         end
     - name: Order_Priority
       expr: split(o_orderpriority, '-')[1]
     measures:
     - name: count_orders
       expr: count(1)
     - name: total_Revenue
       expr: SUM(o_totalprice)
     - name: total_Revenue_per_Customer
       expr: SUM(o_totalprice) / count(distinct o_custkey)
     - name: Total_Revenue_for_Open_Orders
       expr: SUM(o_totalprice) filter (where o_orderstatus='O')
                                 "
  Language                       YAML
  Table Properties               [metric_view.from.name=samples.tpch.orders, metric_view.from.type=ASSET, metric_view.where=o_orderdate > '1990-01-01']

-- Tracking total_revenue_per_customer by month in 1995
> SELECT extract(month from month) as month,
    measure(total_revenue_per_customer)::bigint AS total_revenue_per_customer
  FROM region_sales_metrics
  WHERE extract(year FROM month) = 1995
  GROUP BY ALL
  ORDER BY ALL;
  month	 total_revenue_per_customer
  -----  --------------------------
   1     167727
   2     166237
   3     167349
   4     167604
   5     166483
   6     167402
   7     167272
   8     167435
   9     166633
  10     167441
  11     167286
  12     167542

-- Tracking total_revenue_per_customer by month and status in 1995
> SELECT extract(month from month) as month,
    status,
    measure(total_revenue_per_customer)::bigint AS total_revenue_per_customer
  FROM region_sales_metrics
  WHERE extract(year FROM month) = 1995
  GROUP BY ALL
  ORDER BY ALL;
  month  status      total_revenue_per_customer
  -----  ---------   --------------------------
   1     Fulfilled   167727
   2     Fulfilled   161720
   2    Open          40203
   2    Processing   193412
   3    Fulfilled    121816
   3    Open          52424
   3    Processing   196304
   4    Fulfilled     80405
   4    Open          75630
   4    Processing   196136
   5    Fulfilled     53460
   5    Open         115344
   5    Processing   196147
   6    Fulfilled     42479
   6    Open         160390
   6    Processing   193461
   7    Open         167272
   8    Open         167435
   9    Open         166633
   10   Open         167441
   11   Open         167286
   12   Open         167542