Table Samplers

Postgres provides support for custom tablesamplers as part of the SELECT statement.

ZomboDB provides three different table sampler implementations that allow you to sample a table using a simple text-search query, Elasticsearch's "sampler" aggregate and "diversified_sampler" aggregations.

Sampler Functions

TABLESAMPLE zdb.query_sampler(index_name regclass, query zdbquery)

This is the simplest table sampler that simply returns all the rows that match the specified query. Its results are stable per run, regardless of the REPEATABLE(seed) option.

TABLESAMPLE zdb.sampler(index_name regclass, shard_size int, query zdbquery)

This table sampler uses Elasticsearch's sampler aggregation to sample the specified index using a specific shard size and query.

Per the Elasticsearch documentation, a few example uses cases for this sampler might be

• Tightening the focus of analytics to high-relevance matches rather than the potentially very long tail of low-quality matches
• Reducing the running cost of aggregations that can produce useful results using only samples (e.g. zdb.significant_terms() aggregation)

An example usage might be:

SELECT * 
  FROM table 
  TABLESAMPLE zdb.sampler('idxtable', 200, 'tags:postgres OR tags:sql');

The table sampler can, of course, be combined with a WHERE clause so that you're then searching the results of the sample:

SELECT * 
  FROM table 
  TABLESAMPLE zdb.sampler('idxtable', 200, 'tags:postgres OR tags:sql')
  WHERE table ==> 'zombodb';

The results are stable per run, regardless of the REPEATABLE(seed) option.

TABLESAMPLE zdb.diversified_sampler(index_name regclass, shard_size int, common_field_name text, query zdbquery)

This table sampler uses Elasticsearch's diversified sampler aggregation to sample the specified index.

Per the Elasticsearch docs, this sampler can be used in siutations where you want to

• Tighten the focus of analytics to high-relevance matches rather than the potentially very long tail of low-quality matches
• Remove bias from analytics by ensuring fair representation of content from different sources
• Reduce the running cost of aggregations that can produce useful results using only samples (e.g. zdb.significant_terms() aggregation)

An example usage might be:

SELECT * 
  FROM table 
  TABLESAMPLE zdb.diversified_sampler('idxtable', 200, 'author', 'tags:postgres OR tags:sql');

The table sampler can, of course, be combined with a WHERE clause so that you're then searching the results of the sample:

SELECT * 
  FROM table 
  TABLESAMPLE zdb.diversified_sampler('idxtable', 200, 'author', 'tags:postgres OR tags:sql')
  WHERE table ==> 'zombodb';

The results are stable per run, regardless of the REPEATABLE(seed) option.