Used in: gor only

PARTGOR¶

The PARTGOR command, like PGOR, can be used to run queries in parallel. Unlike the PGOR command which partitions the query execution along the genomic axis, PARTGOR allows for parallel execution of queries along the tag-partition axis of a GOR dictionary table.

The query expression must be provided like a nested-query expression with the variable #{tags} to control filtering on the GOR command as shown in the examples. NOTE: if #{tags} is not provided, the query will fail.

The variable #{tags} can be extended with :q and :dq options. Example #{tags:q} for single quoted tags and #{tags:dq} for double quoted tags.

Like PGOR, PARTGOR cannot appear in a regular nested query.

Usage¶

create #tempfile# = partgor -dict dictionarytable [attributes] <(queryexpression with #{tags} );
gor [#tempfile#]

Options¶

`-parts`	Overrides the default partition split derived from the dictionary table.
`-partsize`	Define the maximum number of tags per part.
`-partscale`	Define the maximum number of tags per part as a ratio of the maximum bucket size.
`-f`	Filter tag list (see the basic GOR command).
`-ff`	File filter relation (see the basic GOR command). Note that in current implementation, this relation CANNOT refer to a temporary table generated with a CREATE statement.

Examples¶

A full discussion of PARTGOR commands and examples of how to use them can be found in the manual chapter on Parallelization in GOR.