Flight Data — On-Time Performance Trend Over Years

Flight Data benchmark · all Flight Data queries

Annual on-time arrival percentage trend

Representative SQL

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

SELECT
  f.year,
  COUNT(*) AS total_flights,
  SUM(CASE WHEN f.cancelled = 1 THEN 1 ELSE 0 END) AS cancelled_flights,
  ROUND(100.0 * SUM(CASE WHEN f.cancelled = 1 THEN 1 ELSE 0 END) / COUNT(*), 2) AS cancellation_rate_pct,
  SUM(CASE WHEN f.cancelled = 0 AND f.arr_delay <= 15 THEN 1 ELSE 0 END) AS ontime_flights,
  ROUND(
    100.0 * SUM(CASE WHEN f.cancelled = 0 AND f.arr_delay <= 15 THEN 1 ELSE 0 END) / NULLIF(SUM(CASE WHEN f.cancelled = 0 THEN 1 ELSE 0 END), 0),
    2
  ) AS ontime_pct,
  ROUND(AVG(CASE WHEN f.cancelled = 0 THEN f.arr_delay END), 2) AS avg_arr_delay
FROM flights AS f
WHERE
  f.flight_date >= '2021-08-01' AND f.flight_date < '2025-01-01'
GROUP BY
  f.year
ORDER BY
  f.year

Representative DataFrame

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

def improvement_trend_expression_impl(ctx: DataFrameContext) -> Any:
    flights, col, lit = _date_window(ctx)

    return (
        flights.with_columns(
            (col("cancelled") == lit(1)).cast(int).alias("_cancelled"),
            ((col("cancelled") == lit(0)) & (col("arr_delay") <= lit(15))).cast(int).alias("_ontime"),
            (col("cancelled") == lit(0)).cast(int).alias("_operated"),
            ctx.when(col("cancelled") == lit(0)).then(col("arr_delay")).otherwise(lit(None)).alias("_arr_nc"),
        )
        .group_by("year")
        .agg(
            col("flight_id").count().alias("total_flights"),
            col("_cancelled").sum().alias("cancelled_flights"),
            col("_ontime").sum().alias("ontime_flights"),
            col("_operated").sum().alias("_non_cancelled"),
            col("_arr_nc").mean().alias("_avg_arr"),
        )
        .with_columns(
            ctx.when(col("_non_cancelled") > lit(0))
            .then(col("_non_cancelled"))
            .otherwise(lit(None))
            .alias("_non_cancelled_nz"),
        )
        .with_columns(
            _round_half_away(ctx, lit(100.0) * col("cancelled_flights") / col("total_flights"), 2).alias(
                "cancellation_rate_pct"
            ),
            _round_half_away(ctx, lit(100.0) * col("ontime_flights") / col("_non_cancelled_nz"), 2).alias("ontime_pct"),
            _round_half_away(ctx, col("_avg_arr"), 2).alias("avg_arr_delay"),
        )
        .select(
            "year",
            "total_flights",
            "cancelled_flights",
            "cancellation_rate_pct",
            "ontime_flights",
            "ontime_pct",
            "avg_arr_delay",
        )
        .sort("year")
    )

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()['improvement-trend'])

For the DataFrame form:

from benchbox.core.query_catalog import get_dataframe_render

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

Source

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