First Notebook

Most medterm4ds work is expected to happen in Python notebooks. Keep one connection open for the notebook session, then use typed results or DataFrames depending on the task.

import medterm4ds as mt

terms = mt.connect("/mnt/d/medterm4ds/data/umls_current.duckdb", memory_profile="low")

Search terminology before choosing codes:

terms.search_df("diabetes", sources=["ICD10CM", "SNOMEDCT_US"], limit=10)

Review exact lookup and patient-friendly names:

codes = ["E11.9", "E11.40", "E11.42"]

lookup = terms.lookup_df("ICD10CM", codes)
friendly = terms.patient_friendly_df("ICD10CM", codes)

friendly[["source", "code", "name", "match_type", "match_depth", "matched_via"]]

Map codes and keep provenance visible:

mapping = terms.map_df(
    "ICD10CM",
    codes,
    target_sources=["SNOMEDCT_US"],
    max_depth=2,
)

mapping[
    [
        "source",
        "code",
        "target_source",
        "target_code",
        "target_display",
        "match_type",
        "match_depth",
        "matched_via",
    ]
]

Optimize a value set:

optimized = terms.optimize("ICD10CM", codes)
optimized.to_dict()

Close the connection when the notebook is finished:

terms.close()