Skip to content

Data retrieval

Match

Given a primary key and a source dataset, retrieves all primary keys that share its cluster in both the source and target datasets. Useful for making ad-hoc queries about specific items of data.

import matchbox as mb
from matchbox import select
import sqlalchemy

engine = sqlalchemy.create_engine('postgresql://')

mb.match(
    select("datahub_companies", engine=engine),
    source=select("companies_house", engine=engine),
    source_pk="8534735",
    resolution_name="last_linker",
)

[
    {
        "cluster": 2354,
        "source": "dbt.companieshouse",
        "source_id": ["8534735", "8534736"],
        "target": "hmrc.exporters",
        "target_id": ["EXP123", "EXP124"]
    }
]

Query

Retrieves entire data sources along with a unique entity identifier according to a point of resolution.

Use Cases

  • Large-scale statistical analysis
  • Building linking or deduplication pipelines
import matchbox as mb
from matchbox import select
import sqlalchemy

engine = sqlalchemy.create_engine('postgresql://')

mb.query(
    select(
        {
            "dbt.companieshouse": ["company_name"],
            "hmrc.exporters": ["year", "commodity_codes"],
        },
        engine=engine,
        combine_type="explode",
        resolution="companies",
    )
)

id      dbt_companieshouse_company_name         hmrc_exporters_year     hmrc_exporters_commodity_codes
122     Acme Ltd.                               2023                    ['85034', '85035']
122     Acme Ltd.                               2024                    ['72142', '72143']
5       Gamma Exports                           2023                    ['90328', '90329']
...

For more information on how to use the functions on this page, please check out the relevant examples in the client API docs.