This R package provides function for evaluating automated concept set constructors.
Concept set construction is the task of creating an OHDSI concept set given a concept set name and possibly a description.
# Evaluating full concept construction using the MMM Challenge gold standard --------
# Using MMM Challenge AI participant 3's submissions:
data(ai3ConceptSets)
results <- evaluateConceptSets(ai3ConceptSets)
results$f1ConservativeWeighted[results$id == "Total"]
# [1] 0.8764369
# Evaluating concept adjudication only ----------------------------------------------
concepts <- getConceptsForAdjudication()
# Using random adjudication for this example:
concepts$adjudication <- sample(c("YES", "NO"), nrow(concepts), replace = TRUE)
evaluateConceptAdjudication(concepts)
# targetDomain tp fp tn fn ppv sensitivity specificity
# 1 CONDITION 59 234 258 49 0.201 0.546 0.524
# 2 MEASUREMENT 38 116 117 29 0.247 0.567 0.502
# 3 PROCEDURE 54 155 160 31 0.258 0.635 0.508
# 4 All 151 505 535 109 0.230 0.581 0.514ConceptSetConstructionEvaluation is an R package. The gold standard is stored as Microsoft Excel files.
See the instructions here for configuring your R environment.
In R, use the following commands to download and install ConceptSetConstructionEvaluation:
install.packages("remotes")
remotes::install_github("ohdsi/ConceptSetConstructionEvaluation")Documentation can be found on the package website.
PDF versions of the documentation are also available:
Read here how you can contribute to this package.