QBVectorColumn

Overview

The QBVectorColumn class represents a vector column in a QBSelect query.

For more information about constructing and executing queries and columns, refer to the QBSelect section of this documentation.

Extends

QBColumn

Properties Summarized

Type	Name	Summary
QBCondition	isNull	Compare column with null.
QBVectorColumn	max	Create an aggregate max expression.
QBVectorColumn	min	Create an aggregate min expression.
QBColumnComparable	not	Create a negated condition.

Methods Summarized

Type	Name	Summary
QBCondition	between(value1, value2)	Compare column to a range of 2 values or other columns.
QBVectorColumn	coalesce(value)	Create coalesce(arg) expression
QBCondition	eq(value)	Compare column with a value or another column.
QBCondition	ge(value)	Compare column with a value or another column.
QBCondition	gt(value)	Compare column with a value or another column.
QBCondition	isin(query)	Compare column with subquery result.
QBCondition	isin(values)	Compare column with values.
QBCondition	isin(customQuery, args)	Compare column with custom query result.
QBCondition	le(value)	Compare column with a value or another column.
QBCondition	lt(value)	Compare column with a value or another column.
QBVectorColumn	nullif(arg)	Create nullif(arg) expression
QBNumberColumn	vector_distance(embedding)	Calculate the database native cosine distance for this column using the embedding.
QBScoreColumn	vector_score(embedding)	Calculate the normalized score for this column using the embedding.

Properties Detailed

isNull

Compare column with null.

Type QBCondition a QBCondition representing the "is null" comparison.

Sample

query.where.add(query.columns.flag.isNull)

max

Create an aggregate max expression.

Type QBVectorColumn a QBColumn representing the sum aggregate function.

Sample

var query = datasources.db.example_data.orders.createSelect();
	query.groupBy.addPk() // have to group by on pk when using having-conditions in (foundset) pk queries
	.root.having.add(query.joins.orders_to_order_details.columns.quantity.count.max(10))
	foundset.loadRecords(query)

min

Create an aggregate min expression.

Type QBVectorColumn a QBColumn representing the minimum aggregate function.

Sample

var query = datasources.db.example_data.orders.createSelect();
	query.groupBy.addPk() // have to group by on pk when using having-conditions in (foundset) pk queries
	.root.having.add(query.joins.orders_to_order_details.columns.quantity.count.min(10))
	foundset.loadRecords(query)

not

Create a negated condition.

Type QBColumnComparable a QBColumn representing the negated condition.

Sample

query.where.add(query.columns.flag.not.eq(1))

Methods Detailed

between(value1, value2)

Compare column to a range of 2 values or other columns.

Parameters

• Object value1 ;

• Object value2 ;

Returns: QBCondition a QBCondition representing the "between" comparison for the two values.

Sample

query.where.add(query.columns.flag.between(0, 5))

coalesce(value)

Create coalesce(arg) expression

Parameters

• Object value when column is null

Returns: QBVectorColumn a QBColumn representing the coalesce expression.

Sample

query.result.add(query.columns.mycol.coalesce('defval'))

eq(value)

Compare column with a value or another column. Operator: equals

Parameters

• Object value ;

Returns: QBCondition a QBCondition representing the "equals" comparison.

Sample

query.where.add(query.columns.flag.eq(1))

ge(value)

Compare column with a value or another column. Operator: greaterThanOrEqual

Parameters

• Object value ;

Returns: QBCondition a QBCondition representing the "greater than or equal to" comparison.

Sample

query.where.add(query.columns.flag.ge(2))

gt(value)

Compare column with a value or another column. Operator: greaterThan

Parameters

• Object value ;

Returns: QBCondition a QBCondition representing the "greater than" comparison.

Sample

query.where.add(query.columns.flag.gt(0))

isin(query)

Compare column with subquery result.

Parameters

• QBPart query subquery

Returns: QBCondition a QBCondition representing the "in" comparison with a subquery.

Sample

query.where.add(query.columns.flag.isin(query2))

isin(values)

Compare column with values.

Parameters

• Array values array of values

Returns: QBCondition a QBCondition representing the "in" comparison with a list of values.

Sample

query.where.add(query.columns.flag.isin([1, 5, 99]))

isin(customQuery, args)

Compare column with custom query result.

Parameters

• String customQuery custom query

• Array args query arguments

Returns: QBCondition a QBCondition representing the "in" comparison with a custom query and arguments.

Sample

query.where.add(query.columns.ccy.isin("select ccycode from currencies c where c.category = " + query.getTableAlias() + ".currency_category and c.flag = ?", ['T']))

le(value)

Compare column with a value or another column. Operator: lessThanOrEqual

Parameters

• Object value ;

Returns: QBCondition a QBCondition representing the "less than or equal to" comparison.

Sample

query.where.add(query.columns.flag.le(2))

lt(value)

Compare column with a value or another column. Operator: lessThan

Parameters

• Object value ;

Returns: QBCondition a QBCondition representing the "less than" comparison.

Sample

query.where.add(query.columns.flag.lt(99))

nullif(arg)

Create nullif(arg) expression

Parameters

• Object arg object to compare

Returns: QBVectorColumn a QBColumn representing the nullif expression.

Sample

query.result.add(query.columns.mycol.nullif('none'))

vector_distance(embedding)

Calculate the database native cosine distance for this column using the embedding.

The native cosine distance is a positive number where lower means better. When sorting on score you can sort on 'vector_distance(embedding) asc' to get the best matches first.

Parameters

• Array embedding embedding object

Returns: QBNumberColumn the QBNumberColumn that can be used for sorting.

Sample

query.sort.add(query.joins.books_to_books_embeddings.columns.embedding.vector_distance(model.embedding('Magic or Fantasy')))

vector_score(embedding)

Calculate the normalized score for this column using the embedding.

The normalized score is a number from 0 to 1 where higher means better. When sorting on score you can sort on 'vector_score(embedding) desc' to get the best matches first. It is more efficient to sort using the native distance function 'vector_distance(embedding) asc' to get the best matches first.

The score result has a function min_score for filtering on the score, this is optimized for filtering compared to the standard number comparison functions.

Parameters

• Array embedding embedding object

Returns: QBScoreColumn the QBScoreColumn that can be added to the result.

Sample

query.result.addquery.joins.books_to_books_embeddings.columns.embedding.vector_score(model.embedding('Magic or Fantasy'))

---