DataFrame.os_query

DataFrame.os_query(query) → DataFrame[source]

Applies an OpenSearch DSL query to the current DataFrame.

Parameters

query:: Dictionary of the OpenSearch DSL query to apply

Returns

opensearch_py_ml.DataFrame:: opensearch_py_ml DataFrame with the query applied

Examples

Apply a `geo-distance query`_ to a dataset with a geo-point column geoip.location.

>>> from tests import OPENSEARCH_TEST_CLIENT

>>> df = oml.DataFrame(OPENSEARCH_TEST_CLIENT, 'ecommerce', columns=['customer_first_name', 'geoip.city_name'])
>>> df.os_query({"bool": {"filter": {"geo_distance": {"distance": "1km", "geoip.location": [55.3, 25.3]}}}}).head()
   customer_first_name geoip.city_name
1                 Mary           Dubai
9            Rabbia Al           Dubai
10           Rabbia Al           Dubai
22                Mary           Dubai
30              Robbie           Dubai

[5 rows x 2 columns]

If using an occurrence like must or filter you must nest it within bool:

# Correct:
df.os_query({
    "bool": {
        "filter": {...}
    }
})

# Incorrect, needs to be nested under 'bool':
df.os_query({
    "filter": {...}
})