atoti.MultiRowArrayConversion

class atoti.MultiRowArrayConversion

Convert an external table where array values are stored with one value per row to a table with array columns.

The external table must have an index_column and at least one “value” column representing the array elements.

All the table columns except from index_column and the array_columns will become key columns.

Example

>>> external_database = session.connect_to_external_database(connection_info)
>>> external_table = external_database.tables["TUTORIAL", "MULTI_ROW_QUANTITY"]

external_table has an INDEX column:

>>> external_table.columns
['PRODUCT', 'INDEX', 'QUANTITY']

and its content is:

PRODUCT	INDEX	QUANTITY
product_1	0	10.0
product_1	1	20.0
product_1	2	15.0
product_1	3	25.0
product_1	4	10.0
product_2	0	50.0
product_2	1	65.0
product_2	2	55.0
product_2	3	30.0
product_2	4	80.0

It can be converted into a table with an array column:

>>> from atoti_directquery_snowflake import SnowflakeTableOptions
>>> table = session.add_external_table(
...     external_table,
...     table_name="Sales (Multi row array)",
...     options=SnowflakeTableOptions(
...         array_conversion=tt.MultiRowArrayConversion(
...             index_column="INDEX",
...             array_columns={"QUANTITY"},
...         ),
...     ),
... )
>>> table.head().sort_index()
                                 QUANTITY
PRODUCT
product_1  [10.0, 20.0, 15.0, 25.0, 10.0]
product_2  [50.0, 65.0, 55.0, 30.0, 80.0]

array_columns: Set[str] | Sequence[str]

Names of the columns that contain array values.

index_column: str

Name of the column used as an index for the arrays.