Skip to content

functions.md

Latest commit

 

History

History
155 lines (119 loc) · 3.99 KB

functions.md

File metadata and controls

155 lines (119 loc) · 3.99 KB

Functions Reference

Aggregate Functions

These functions operate on groups of rows in SELECT queries with GROUP BY.

Function Description Example
COUNT(*) Count all rows SELECT COUNT(*) FROM users
COUNT(column) Count non-null values SELECT COUNT(name) FROM users
SUM(column) Sum numeric values SELECT SUM(total) FROM orders
AVG(column) Average numeric values SELECT AVG(age) FROM users
MIN(column) Minimum value SELECT MIN(created) FROM orders
MAX(column) Maximum value SELECT MAX(total) FROM orders 
SELECT team, COUNT(*) AS headcount, AVG(age) AS avg_age
FROM employees
GROUP BY team
HAVING COUNT(*) > 5
ORDER BY headcount DESC

Graph Algorithm Functions

These are invoked as top-level commands, not as SQL functions. See the Query Language Reference for full syntax.

Algorithm Description Returns
PAGERANK Link analysis ranking Node scores (0.0-1.0)
BETWEENNESS Bridge node importance Node scores
CLOSENESS Average distance to all nodes Node scores
EIGENVECTOR Influence-based ranking Node scores
LOUVAIN Community detection Community assignments
LABEL_PROPAGATION Community detection Community assignments

Graph Aggregate Functions

Used with the GRAPH AGGREGATE command on node/edge properties.

Function Description
COUNT Count nodes/edges matching criteria
SUM Sum of property values
AVG Average of property values
MIN Minimum property value
MAX Maximum property value 
GRAPH AGGREGATE COUNT NODES person
GRAPH AGGREGATE AVG NODE age person WHERE age > 20
GRAPH AGGREGATE SUM EDGE weight collaborates

Distance Metrics

Used with SIMILAR and EMBED commands for vector similarity search.

Metric Keyword Range Best For
Cosine similarity COSINE -1.0 to 1.0 Text embeddings, normalized vectors
Euclidean distance EUCLIDEAN 0.0 to infinity Spatial data, image features
Dot product DOT_PRODUCT -infinity to infinity Pre-normalized vectors, recommendation 
SIMILAR [0.1, 0.2, 0.3] LIMIT 10 METRIC COSINE
SIMILAR 'doc1' LIMIT 5 METRIC EUCLIDEAN

The default metric is COSINE when not specified.

Expression Operators

Arithmetic

Operator Description
+ Addition
- Subtraction
* Multiplication
/ Division
% Modulo

Comparison

Operator Description
= Equal
!= or <> Not equal
< Less than
<= Less than or equal
> Greater than
>= Greater than or equal

Logical

Operator Description
AND Logical AND
OR Logical OR
NOT Logical NOT

Special Predicates

Predicate Description Example
IS NULL Test for null WHERE name IS NULL
IS NOT NULL Test for non-null WHERE name IS NOT NULL
IN (list) Set membership WHERE id IN (1, 2, 3)
NOT IN (list) Set non-membership WHERE id NOT IN (1, 2)
BETWEEN a AND b Range check WHERE age BETWEEN 18 AND 65
LIKE pattern Pattern matching WHERE name LIKE 'A%'
NOT LIKE pattern Negative pattern WHERE name NOT LIKE '%test%'
EXISTS (subquery) Subquery existence WHERE EXISTS (SELECT ...)

CASE Expression

CASE
    WHEN condition THEN result
    [WHEN condition THEN result ...]
    [ELSE default]
END
SELECT name,
    CASE
        WHEN age < 18 THEN 'minor'
        WHEN age < 65 THEN 'adult'
        ELSE 'senior'
    END AS category
FROM users

CAST

CAST(expression AS type)
SELECT CAST(age AS FLOAT) / 10 AS decade FROM users