TPC-DS-OBT Q14 — obt_bulk_sales

TPC-DS-OBT benchmark · all TPC-DS-OBT queries

Multi-unit sales volume and revenue by channel

Representative SQL

Rendered for DataFusion with default parameters.

WITH cross_items AS (
  SELECT
    item_i_item_sk,
    item_i_brand_id,
    item_i_class_id,
    item_i_category_id
  FROM (
    SELECT
      item_i_item_sk,
      item_i_brand_id,
      item_i_class_id,
      item_i_category_id,
      channel
    FROM tpcds_sales_returns_obt
    WHERE
      sold_date_d_year BETWEEN 1998 AND 1998 + 2
    GROUP BY
      item_i_item_sk,
      item_i_brand_id,
      item_i_class_id,
      item_i_category_id,
      channel
  ) AS items_by_channel
  GROUP BY
    item_i_item_sk,
    item_i_brand_id,
    item_i_class_id,
    item_i_category_id
  HAVING
    COUNT(DISTINCT channel) = 3
), avg_sales AS (
  SELECT
    AVG(quantity * list_price) AS average_sales
  FROM tpcds_sales_returns_obt
  WHERE
    sold_date_d_year BETWEEN 1998 AND 1998 + 2
)
SELECT
  channel,
  i_brand_id,
  i_class_id,
  i_category_id,
  SUM(sales),
  SUM(number_sales)
FROM (
  SELECT
    channel,
    item_i_brand_id AS i_brand_id,
    item_i_class_id AS i_class_id,
    item_i_category_id AS i_category_id,
    SUM(quantity * list_price) AS sales,
    COUNT(*) AS number_sales
  FROM tpcds_sales_returns_obt
  WHERE
    item_sk IN (
      SELECT
        item_i_item_sk
      FROM cross_items
    )
    AND sold_date_d_year = 1998 + 2
    AND sold_date_d_moy = 11
  GROUP BY
    channel,
    item_i_brand_id,
    item_i_class_id,
    item_i_category_id
  HAVING
    SUM(quantity * list_price) > (
      SELECT
        average_sales
      FROM avg_sales
    )
) AS y
GROUP BY
  ROLLUP (
    channel,
    i_brand_id,
    i_class_id,
    i_category_id
  )
ORDER BY
  channel,
  i_brand_id,
  i_class_id,
  i_category_id
LIMIT 100

Representative DataFrame

Expression family (Polars, PySpark, DataFusion), default parameters.

def impl(ctx: DataFrameContext) -> Any:
        table = ctx.get_table(OBT_TABLE)
        for column, op, value in spec.filters:
            table = table.filter(_FILTER_OPS[op](ctx.col(column), ctx.lit(value)))
        aggregates = [getattr(ctx.col(column), func)().alias(alias) for alias, column, func in spec.aggregations]
        if spec.group_keys:
            result = table.group_by(*spec.group_keys).agg(*aggregates)
        else:
            result = table.select(*aggregates)
        if spec.sort_keys:
            descending = list(spec.descending) if spec.descending else [False] * len(spec.sort_keys)
            result = result.sort(list(spec.sort_keys), descending=descending)
        if spec.limit:
            result = result.limit(spec.limit)
        return result

Get the exact query for your run

The blocks above are representative. To get the precise statement a specific platform runs at a specific scale, substitute your own values:

from benchbox import TPCDSOBT

bench = TPCDSOBT(scale_factor=1.0)            # your scale factor
queries = bench.get_queries(dialect="datafusion")   # your platform's dialect; drop dialect= for native SQL
print(queries['14'])

For the DataFrame form:

from benchbox.core.query_catalog import get_dataframe_render

print(get_dataframe_render('tpcds_obt', '14', family="expression").source)  # or family="pandas"

Source

Query text: benchbox/core/tpcds_obt/queries.py