Package dev.frostlake.functions.window

Class WindowFunctionHelper

java.lang.Object

dev.frostlake.functions.window.WindowFunctionHelper

public final class WindowFunctionHelper
extends Object

Shared logic for Snowflake window functions. All methods work on pre-sorted partition row lists.

Method Summary

Modifier and Type	Method	Description
static Object	avg(List<Object> values)	AVG over the frame, ignoring nulls; null when there are no non-null values.
static Object	avg(List<Object> values, boolean variantArgument)	See sum(List, boolean) — the same declared-VARIANT rule applied to a window AVG.
static Object	avg(List<List<Object>> partials, boolean variantArgument, boolean nonCumulative, boolean rowWise)	AVG over a window frame given as the partials its FLOAT sum is read in, at a cumulative window's scale or a non-cumulative one's — see AggregateNumerics.windowAvg(List, boolean, boolean, boolean).
static int	compareValues(Object a, Object b)
static double	cumeDist(List<Row> sortedRows, Row currentRow, List<Object> orderValues)	CUME_DIST — fraction of rows with order value <= current; ranges (0, 1].
static Object	firstValue(List<Object> frameValues, boolean ignoreNulls)	FIRST_VALUE(expr) — value of expr in the first row of the window frame.
static int	indexOf(List<Row> rows, Row target)
static Object	lastValue(List<Object> frameValues, boolean ignoreNulls)	LAST_VALUE(expr) — value of expr in the last row of the window frame.
static Object	max(List<Object> values)	MAX over the frame (by compareValues(java.lang.Object, java.lang.Object)), ignoring nulls; preserves the element's type.
static Object	min(List<Object> values)	MIN over the frame (by compareValues(java.lang.Object, java.lang.Object)), ignoring nulls; preserves the element's type.
static Object	nonCumulativeAvg(List<Object> values, boolean variantArgument)	AVG over a NON-CUMULATIVE window, which declares three decimals wider than its input rather than six and TRUNCATES to that scale — see AggregateNumerics.nonCumulativeWindowAvg(Iterable, boolean).
static Object	nthValue(List<Object> frameValues, int n)	NTH_VALUE(expr, n) — value of expr in the nth row of the partition (1-based).
static long	ntile(List<Row> sortedRows, Row currentRow, int buckets)	NTILE(n) — divides the sorted partition into n equal buckets, returns bucket number (1-based).
static double	percentRank(List<Row> sortedRows, Row currentRow, List<Object> orderValues)	PERCENT_RANK — (rank - 1) / (N - 1); ranges [0, 1].
static Object	ratioToReport(Object currentValue, List<Object> allValues)	RATIO_TO_REPORT(expr) — value / SUM(value over partition).
static Object	sum(List<Object> values)	SUM over the frame, ignoring nulls; null when the frame has no non-null values (SQL SUM semantics).
static Object	sum(List<Object> values, boolean variantArgument)	The declared-VARIANT-argument flavour — live, SUM(v:b) OVER () is DOUBLE (SYSTEM$TYPEOF FLOAT) exactly like the grouped form, even over whole-number JSON values.
static Object	sum(List<List<Object>> partials, boolean variantArgument, boolean rowWise)	SUM over a window frame given as the partials its FLOAT sum is read in — see AggregateNumerics.windowSum(List, boolean, boolean).
static double	toDouble(Object v)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

ntile

public static long ntile(List<Row> sortedRows,
 Row currentRow,
 int buckets)

NTILE(n) — divides the sorted partition into n equal buckets, returns bucket number (1-based).

percentRank

public static double percentRank(List<Row> sortedRows,
 Row currentRow,
 List<Object> orderValues)

PERCENT_RANK — (rank - 1) / (N - 1); ranges [0, 1].

cumeDist

public static double cumeDist(List<Row> sortedRows,
 Row currentRow,
 List<Object> orderValues)

CUME_DIST — fraction of rows with order value <= current; ranges (0, 1].

ratioToReport

public static Object ratioToReport(Object currentValue,
 List<Object> allValues)

RATIO_TO_REPORT(expr) — value / SUM(value over partition).

firstValue

public static Object firstValue(List<Object> frameValues,
 boolean ignoreNulls)

FIRST_VALUE(expr) — value of expr in the first row of the window frame. With RESPECT NULLS (Snowflake's default) the first row's value is returned even if it is NULL; with IGNORE NULLS the first non-null value is returned (null if the frame has none).

lastValue

public static Object lastValue(List<Object> frameValues,
 boolean ignoreNulls)

LAST_VALUE(expr) — value of expr in the last row of the window frame. RESPECT NULLS (default) returns the last row's value even if NULL; IGNORE NULLS returns the last non-null value.

nthValue

public static Object nthValue(List<Object> frameValues,
 int n)

NTH_VALUE(expr, n) — value of expr in the nth row of the partition (1-based).

indexOf

public static int indexOf(List<Row> rows,
 Row target)

compareValues

public static int compareValues(Object a,
 Object b)

toDouble

public static double toDouble(Object v)

sum

public static Object sum(List<Object> values)

SUM over the frame, ignoring nulls; null when the frame has no non-null values (SQL SUM semantics).

sum

public static Object sum(List<Object> values,
 boolean variantArgument)

The declared-VARIANT-argument flavour — live, SUM(v:b) OVER () is DOUBLE (SYSTEM$TYPEOF FLOAT) exactly like the grouped form, even over whole-number JSON values.

avg

public static Object avg(List<Object> values)

AVG over the frame, ignoring nulls; null when there are no non-null values. Shares the grouped AVG's Snowflake typing: fixed-point inputs average to a scale-(max input scale + 6) BigDecimal, any double/VARIANT input keeps the double average.

avg

public static Object avg(List<Object> values,
 boolean variantArgument)

See sum(List, boolean) — the same declared-VARIANT rule applied to a window AVG.

nonCumulativeAvg

public static Object nonCumulativeAvg(List<Object> values,
 boolean variantArgument)

AVG over a NON-CUMULATIVE window, which declares three decimals wider than its input rather than six and TRUNCATES to that scale — see AggregateNumerics.nonCumulativeWindowAvg(Iterable, boolean).

sum

public static Object sum(List<List<Object>> partials,
 boolean variantArgument,
 boolean rowWise)

SUM over a window frame given as the partials its FLOAT sum is read in — see AggregateNumerics.windowSum(List, boolean, boolean).

avg

public static Object avg(List<List<Object>> partials,
 boolean variantArgument,
 boolean nonCumulative,
 boolean rowWise)

AVG over a window frame given as the partials its FLOAT sum is read in, at a cumulative window's scale or a non-cumulative one's — see AggregateNumerics.windowAvg(List, boolean, boolean, boolean).

min

public static Object min(List<Object> values)

MIN over the frame (by compareValues(java.lang.Object, java.lang.Object)), ignoring nulls; preserves the element's type.

max

public static Object max(List<Object> values)

MAX over the frame (by compareValues(java.lang.Object, java.lang.Object)), ignoring nulls; preserves the element's type.