Flight Data — Late Aircraft (Cascade) Delay Patterns

Flight Data benchmark · all Flight Data queries

Flights delayed due to late incoming aircraft by carrier

Representative SQL

Rendered in the benchmark’s native SQL with default parameters (no dialect translation available).

SELECT
  f.reporting_airline,
  a.name AS airline_name,
  COUNT(*) AS total_flights,
  SUM(CASE WHEN f.late_aircraft_delay > 0 THEN 1 ELSE 0 END) AS cascade_delayed,
  ROUND(100.0 * SUM(CASE WHEN f.late_aircraft_delay > 0 THEN 1 ELSE 0 END) / COUNT(*), 2) AS cascade_rate_pct,
  ROUND(AVG(CASE WHEN f.late_aircraft_delay > 0 THEN f.late_aircraft_delay END), 2) AS avg_cascade_delay,
  FLOOR(
    (
      SUM(
        CAST(CASE
          WHEN f.late_aircraft_delay > 0
          THEN ROUND(f.late_aircraft_delay * 10)
          ELSE 0
        END AS BIGINT)
      ) + 5
    ) / 10
  ) AS total_cascade_minutes
FROM flights AS f
LEFT JOIN airlines AS a
  ON f.reporting_airline = a.code
WHERE
  f.cancelled = 0 AND f.flight_date >= '2021-08-01' AND f.flight_date < '2025-01-01'
GROUP BY
  f.reporting_airline,
  a.name
ORDER BY
  cascade_rate_pct DESC,
  f.reporting_airline

Representative DataFrame

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

def cascade_delays_expression_impl(ctx: DataFrameContext) -> Any:
    flights = ctx.get_table("flights")
    col, lit = ctx.col, ctx.lit

    return (
        flights.join(_airline_names(ctx), left_on="reporting_airline", right_on="a_code", how="left")
        .filter(_date_condition(col, lit, lambda col, lit: col("cancelled") == lit(0)))
        .with_columns(
            (col("late_aircraft_delay") > lit(0)).cast(int).alias("_cascade"),
            ctx.when(col("late_aircraft_delay") > lit(0))
            .then(col("late_aircraft_delay"))
            .otherwise(lit(None))
            .alias("_cascade_pos"),
            ctx.when(col("late_aircraft_delay") > lit(0))
            .then(_round_half_away(ctx, col("late_aircraft_delay") * lit(10), 0).cast(int))
            .otherwise(lit(0))
            .alias("_late_tenths"),
        )
        .group_by("reporting_airline", "airline_name")
        .agg(
            col("flight_id").count().alias("total_flights"),
            col("_cascade").sum().alias("cascade_delayed"),
            col("_cascade_pos").mean().alias("_avg_cascade"),
            col("_late_tenths").sum().alias("_total_tenths"),
        )
        .with_columns(
            _round_half_away(ctx, lit(100.0) * col("cascade_delayed") / col("total_flights"), 2).alias(
                "cascade_rate_pct"
            ),
            _round_half_away(ctx, col("_avg_cascade"), 2).alias("avg_cascade_delay"),
            ((col("_total_tenths") + lit(5)) / lit(10)).floor().alias("total_cascade_minutes"),
        )
        .select(
            "reporting_airline",
            "airline_name",
            "total_flights",
            "cascade_delayed",
            "cascade_rate_pct",
            "avg_cascade_delay",
            "total_cascade_minutes",
        )
        .sort(["cascade_rate_pct", "reporting_airline"], descending=[True, False])
    )

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 FlightData

bench = FlightData(scale_factor=1.0)            # your scale factor
print(bench.get_queries()['cascade-delays'])

For the DataFrame form:

from benchbox.core.query_catalog import get_dataframe_render

print(get_dataframe_render('flightdata', 'cascade-delays', family="expression").source)  # or family="pandas"

Source

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