Categories:: Aggregate functions (General)

SKEW¶

Returns the sample skewness of non-NULL records. If all records inside a group are NULL, the function returns NULL.

The following formula is used to compute the sample skewness:

\[(n^2)/((n-1) * (n-2)) * (m_3/(k_2)^(1.5))\]

where:

\(n\) denotes the number of non-null records.
\(m_3\) denotes the sample third central moment.
\(k_2\) denotes the symmetric unbiased estimator of the variance.

Intuitively, skew describes how asymmetric the underlying distribution is.

Syntax¶

SKEW( <expr> )

Arguments¶

expr: This is an expression that evaluates to a numeric data type (INTEGER, FLOAT, DECIMAL, etc.).

Returns¶

This function returns a value of type DOUBLE.

Usage notes¶

For inputs with fewer than three records, SKEW returns NULL.

Examples¶

Create a table and load the data:

create or replace table aggr(k int, v decimal(10,2), v2 decimal(10, 2));

insert into aggr values
    (1, 10, null),
    (2, 10, null),
    (2, 20, 22),
    (2, 25, null),
    (2, 30, 35);

Display the data:

select * 
    from aggr
    order by k, v;
+---+-------+-------+
| K |     V |    V2 |
|---+-------+-------|
| 1 | 10.00 |  NULL |
| 2 | 10.00 |  NULL |
| 2 | 20.00 | 22.00 |
| 2 | 25.00 |  NULL |
| 2 | 30.00 | 35.00 |
+---+-------+-------+

Query the data:

select SKEW(K), SKEW(V), SKEW(V2) 
    from aggr;
+--------------+---------------+----------+
|      SKEW(K) |       SKEW(V) | SKEW(V2) |
|--------------+---------------+----------|
| -2.236069766 | 0.05240788515 |     NULL |
+--------------+---------------+----------+