CoffeeShop — Day-Part Cadence

CoffeeShop benchmark · all CoffeeShop queries

Order cadence by day-part (Morning/Midday/Afternoon/Evening) for a region

Representative SQL

Rendered for DataFusion with default parameters.

SELECT
  CASE
    WHEN EXTRACT(HOUR FROM ol.order_time) BETWEEN 5 AND 10
    THEN 'Morning'
    WHEN EXTRACT(HOUR FROM ol.order_time) BETWEEN 11 AND 14
    THEN 'Midday'
    WHEN EXTRACT(HOUR FROM ol.order_time) BETWEEN 15 AND 17
    THEN 'Afternoon'
    ELSE 'Evening'
  END AS day_part,
  COUNT(DISTINCT ol.order_id) AS orders,
  SUM(ol.total_price) AS revenue,
  AVG(ol.total_price) AS avg_line_value
FROM order_lines AS ol
JOIN dim_locations AS dl
  ON ol.location_record_id = dl.record_id
WHERE
  dl.region = 'South'
  AND ol.order_date BETWEEN CAST('2023-01-01' AS DATE) AND CAST('2024-12-31' AS DATE)
GROUP BY
  day_part
ORDER BY
  day_part

Representative DataFrame

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

def tm1_expression_impl(ctx: DataFrameContext) -> Any:
    params = get_parameters("TM1")
    region = params.get("region", "South")
    start_date = params.get("start_date", "2023-01-01")
    end_date = params.get("end_date", "2024-12-31")

    ol = ctx.get_table("order_lines")
    dl = ctx.get_table("dim_locations")
    col = ctx.col
    lit = ctx.lit

    joined = (
        ol.filter(
            (col("order_date") >= lit(_parse_date(start_date))) & (col("order_date") <= lit(_parse_date(end_date)))
        )
        .join(dl, left_on="location_record_id", right_on="record_id")
        .filter(col("region") == lit(region))
    )
    if _is_temporal_expression_column(joined, "order_time"):
        hour_expr = col("order_time").dt.hour()
    else:
        hour_expr = col("order_time").str.slice(0, 2).cast_int()
    with_hour = joined.with_columns(hour_expr.alias("hour"))
    with_part = with_hour.with_columns(
        ctx.when((col("hour") >= lit(5)) & (col("hour") <= lit(10)))
        .then(lit("Morning"))
        .otherwise(
            ctx.when((col("hour") >= lit(11)) & (col("hour") <= lit(14)))
            .then(lit("Midday"))
            .otherwise(
                ctx.when((col("hour") >= lit(15)) & (col("hour") <= lit(17)))
                .then(lit("Afternoon"))
                .otherwise(lit("Evening"))
            )
        )
        .alias("day_part")
    )
    return (
        with_part.group_by("day_part")
        .agg(
            col("order_id").n_unique().alias("orders"),
            col("total_price").sum().alias("revenue"),
            col("total_price").mean().alias("avg_line_value"),
        )
        .sort("day_part")
    )

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 CoffeeShop

bench = CoffeeShop(scale_factor=0.001)            # your scale factor
queries = bench.get_queries(dialect="datafusion")   # your platform's dialect; drop dialect= for native SQL
print(queries['TM1'])

For the DataFrame form:

from benchbox.core.query_catalog import get_dataframe_render

print(get_dataframe_render('coffeeshop', 'TM1', family="expression").source)  # or family="pandas"

Source

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