query: str | Selectable,
connection: ConnectionOrCursor | str,
iter_batches: bool = False,
batch_size: int | None = None,
schema_overrides: SchemaDict | None = None,
infer_schema_length: int | None = 100,
execute_options: dict[str, Any] | None = None,
**kwargs: Any,
) DataFrame | Iterable[DataFrame][source]#

Read the results of a SQL query into a DataFrame, given a connection object.


SQL query to execute (if using a SQLAlchemy connection object this can be a suitable “Selectable”, otherwise it is expected to be a string).


An instantiated connection (or cursor/client object) that the query can be executed against. Can also pass a valid ODBC connection string, identified as such if it contains the string “Driver={…}”, in which case the arrow-odbc package will be used to establish the connection and return Arrow-native data to Polars. Async driver connections are also supported, though this is currently considered unstable.


Use of asynchronous connections is currently considered unstable, and unexpected issues may arise; if this happens, please report them.


Return an iterator of DataFrames, where each DataFrame represents a batch of data returned by the query; this can be useful for processing large resultsets in a memory-efficient manner. If supported by the backend, this value is passed to the underlying query execution method (note that very low values will typically result in poor performance as it will result in many round-trips to the database as the data is returned). If the backend does not support changing the batch size then a single DataFrame is yielded from the iterator.


Indicate the size of each batch when iter_batches is True (note that you can still set this when iter_batches is False, in which case the resulting DataFrame is constructed internally using batched return before being returned to you. Note that some backends may support batched operation but not allow for an explicit size; in this case you will still receive batches, but their exact size will be determined by the backend (so may not equal the value set here).


A dictionary mapping column names to dtypes, used to override the schema inferred from the query cursor or given by the incoming Arrow data (depending on driver/backend). This can be useful if the given types can be more precisely defined (for example, if you know that a given column can be declared as u32 instead of i64).


The maximum number of rows to scan for schema inference. If set to None, the full data may be scanned (this can be slow). This parameter only applies if the data is read as a sequence of rows and the schema_overrides parameter is not set for the given column; Arrow-aware drivers also ignore this value.


These options will be passed through into the underlying query execution method as kwargs. In the case of connections made using an ODBC string (which use arrow-odbc) these options are passed to the read_arrow_batches_from_odbc method.

See also


Create a DataFrame from a SQL query using a URI string.


  • This function supports a wide range of native database drivers (ranging from local databases such as SQLite to large cloud databases such as Snowflake), as well as generic libraries such as ADBC, SQLAlchemy and various flavours of ODBC. If the backend supports returning Arrow data directly then this facility will be used to efficiently instantiate the DataFrame; otherwise, the DataFrame is initialised from row-wise data.

  • Support for Arrow Flight SQL data is available via the adbc-driver-flightsql package; see https://arrow.apache.org/adbc/current/driver/flight_sql.html for more details about using this driver (notable databases implementing Flight SQL include Dremio and InfluxDB).

  • The read_database_uri function can be noticeably faster than read_database if you are using a SQLAlchemy or DBAPI2 connection, as connectorx and adbc optimises translation of the result set into Arrow format. Note that you can determine a connection’s URI from a SQLAlchemy engine object by calling conn.engine.url.render_as_string(hide_password=False).

  • If Polars has to create a cursor from your connection in order to execute the query then that cursor will be automatically closed when the query completes; however, Polars will never close any other open connection or cursor.

  • We are able to support more than just relational databases and SQL queries through this function. For example, we can load graph database results from a KùzuDB connection in conjunction with a Cypher query.


Instantiate a DataFrame from a SQL query against a user-supplied connection:

>>> df = pl.read_database(
...     query="SELECT * FROM test_data",
...     connection=user_conn,
...     schema_overrides={"normalised_score": pl.UInt8},
... )  

Use a parameterised SQLAlchemy query, passing named values via execute_options:

>>> df = pl.read_database(
...     query="SELECT * FROM test_data WHERE metric > :value",
...     connection=alchemy_conn,
...     execute_options={"parameters": {"value": 0}},
... )  

Use ‘qmark’ style parameterisation; values are still passed via execute_options, but in this case the “parameters” value is a sequence of literals, not a dict:

>>> df = pl.read_database(
...     query="SELECT * FROM test_data WHERE metric > ?",
...     connection=alchemy_conn,
...     execute_options={"parameters": [0]},
... )  

Instantiate a DataFrame using an ODBC connection string (requires the arrow-odbc package) setting upper limits on the buffer size of variadic text/binary columns, returning the result as an iterator over DataFrames that each contain 1000 rows:

>>> for df in pl.read_database(
...     query="SELECT * FROM test_data",
...     connection="Driver={PostgreSQL};Server=localhost;Port=5432;Database=test;Uid=usr;Pwd=",
...     execute_options={"max_text_size": 512, "max_binary_size": 1024},
...     iter_batches=True,
...     batch_size=1000,
... ):
...     do_something(df)  

Load graph database results from a KùzuDB connection and a Cypher query:

>>> df = pl.read_database(
...     query="MATCH (a:User)-[f:Follows]->(b:User) RETURN a.name, f.since, b.name",
...     connection=kuzu_db_conn,
... )  

Load data from an asynchronous SQLAlchemy driver/engine; note that asynchronous connections and sessions are also supported here:

>>> from sqlalchemy.ext.asyncio import create_async_engine
>>> async_engine = create_async_engine("sqlite+aiosqlite:///test.db")
>>> df = pl.read_database(
...     query="SELECT * FROM test_data",
...     connection=async_engine,
... )  

Load data from an asynchronous SurrealDB client connection object; note that both the WS (Surreal) and HTTP (SurrealHTTP) clients are supported:

>>> import asyncio
>>> async def surreal_query_to_frame(query: str, url: str):
...     async with Surreal(url) as client:
...         await client.use(namespace="test", database="test")
...         return pl.read_database(query=query, connection=client)
>>> df = asyncio.run(
...     surreal_query_to_frame(
...         query="SELECT * FROM test_data",
...         url="ws://localhost:8000/rpc",
...     )
... )