getAnchors: This function allows Anchor recommendation

In OHDSI/Aphrodite: Automated PHenotype Routine for Observational Definition Identification Training and Evaluation (APHRODITE) - Phenotype building tool using Fuzzy labels

Description

This function allows Anchor recommendation based after your initial set of keyword and ignore lists have been provided. This will help improve model by suggesting related features that were not considered initially.

Usage

getAnchors(connection, dbms, schema, casesList, controlsList, ignores,
  studyName, outcomeName, flag, numAnchors)

Arguments

connection	The connection to the database server.
dbms	The target DBMS for SQL to be rendered in.
schema	The database schema being used.
casesList	The list of case patients (already filtered by keywords or gold standard).
controlsList	The list of control patients.
ignores	The list of concept_id's ignored when building the cohort.
studyName	The study name(will be used for file naming).
outcomeName	The outcomeName (will be use for modeling).
flag	The flags variable containg the study configuration - we use this one here to have flexibility of having two different sets of settings for the same experimental run.
numAnchors	The total number of anchors to be returned (top N features).

Details

This function takes the lists of exclude keywords and fetches all patient data for the patients on the cases and controls list. It then builds a model to identify the top performing features and returns a list of them as anchors. This new keyword list can be feed to the set of Anchors specific functions to use any anchor as a selection criteria for patients

Value

A list of anchors containing rank, conceptID, domaiID

Examples

## Not run: 

numAnchors<-50
anchor_list <- getAnchors(conn, dbms, cdmSchema, cases, controls, as.character(ignoreList_FF$V3), studyName, outcomeName, flag, numAnchors)


## End(Not run)

OHDSI/Aphrodite documentation built on Sept. 18, 2020, 10:26 p.m.