medcat.utils.cdb_utils

Module Contents

Functions

merge_cdb(cdb1, cdb2[, overwrite_training, full_build])

Merge two CDB's together to produce a new, single CDB. The contents of inputs CDBs will not be changed.

Attributes

logger

medcat.utils.cdb_utils.logger
medcat.utils.cdb_utils.merge_cdb(cdb1, cdb2, overwrite_training=0, full_build=False)

Merge two CDB’s together to produce a new, single CDB. The contents of inputs CDBs will not be changed. addl_info can not be perfectly merged, and will prioritise cdb1. see full_build

Parameters:

  • cdb1 (CDB) – The first medcat cdb to merge. In cases where merging isn’t suitable isn’t ideal (such as cui2preferred_name), this cdb values will be prioritised over cdb2.

  • cdb2 (CDB) – The second medcat cdb to merge.

  • overwrite_training (int) – Choose to prioritise a CDB’s context vectors values over merging gracefully. 0 - no prio, 1 - CDB1, 2 - CDB2

  • full_build (bool) – Add additional information from “addl_info” dicts “cui2ontologies” and “cui2description”

Returns:

CDB – The merged CDB.

Return type:

medcat.cdb.CDB