Categories:

Aggregate functions (Linear Regression) , Window functions

REGR_AVGX

Returns the average of the independent variable for non-null pairs in a group, where x is the independent variable and y is the dependent variable.

Syntax

Aggregate function

REGR_AVGX(y, x)

Window function

REGR_AVGX(y, x) OVER ( [ PARTITION BY <expr3> ] )

Arguments

y

The dependent variable. This must be an expression that can be evaluated to a numeric type.

x

The independent variable. This must be an expression that can be evaluated to a numeric type.

expr3

This is the optional expression used to group rows into partitions.

Important

Note the order of the arguments; the dependent variable is first.

Returns

If any of the input expressions is of type DECFLOAT, the returned type is DECFLOAT. Otherwise, the returned type is FLOAT.

Usage notes

  • DISTINCT is not supported for this function.
  • In order for a row to be included in the average, BOTH the x and y values must be non-NULL.
  • When this function is called as a window function, it does not support:
    • An ORDER BY clause within the OVER clause.
    • Explicit window frames.

Examples

CREATE OR REPLACE TABLE aggr(k int, v decimal(10,2), v2 decimal(10, 2));
INSERT INTO aggr VALUES(1, 10, NULL);
INSERT INTO aggr VALUES(2, 10, 11), (2, 20, 22), (2, 25, NULL), (2, 30, 35);
SELECT k, REGR_AVGX(v, v2) FROM aggr GROUP BY k;

---+------------------+
 k | regr_avgx(v, v2) |
---+------------------+
 1 | [NULL]           |
 2 | 22.666666667     |
---+------------------+