dataExtract: This extracts the history features for each person in the...

In jreps/patientCluster: Performs patient clustering for patient cohorts using OHDSI CDM v5

Description

This function connects to the CDM and constructs the data used to do the clustering - this is either condition_concept_ids that are recorded during the defined time period relative to the cohort start date for each person in the cohort or covariate concept_sets that are specified by using the 'default' grouping or inputing a dataframe with columns: definition and concept_id specifiying the concept_ids that make up each covariate definition, see examples below.

Usage

dataExtract(dbconnection = NULL, cdmDatabaseSchema = NULL,
  cohortDatabaseSchema = NULL, cohortid = 100, ageMin = NULL,
  ageMax = NULL, gender = NULL, type = "group", groupDef = "default",
  historyStart = 1, historyEnd = 180, ffloc = NULL, debug = NULL, ...)

Arguments

dbconnection	class:connectionDetails - the database connection details requires Library(DatabaseConnector)
cdmDatabaseSchema	class:character - database schema containing cdm tables
cohortDatabaseSchema	class:character - database schema containing cohort
cohortid	class:numeric - id of cohort in cohort table
gender	class:numeric - gender concept_id (8507- male; 8532-female)
type	class:character - features used by clustering (condition i.e. all condition_concept_ids or group i.e. concept sets),
groupDef	class:dataframe - a dataframe containing covariate concept_sets - must have the columns definition and concept_id
historyStart	class:numeric days prior to index to start searching person records for features
historyEnd	class:numeric days prior to index to stop searching person records for features
ffloc	class:character - specifies the directory where the ff files are stored
debug	class:character - default(NULL) otherise specifies the directory where the main SQL for extraction is written to for debugging
minAge	class:numeric default(NULL)- the minimum age a person in the cohort must be to be included in the data
maxAge	class:numeric default(NULL)- the maximum age a person in the cohort must be to be included in the data

Value

clusterData class:clusterData - a list containing:

strata	an ffdf containing the age/gender/row_id for each person in the cohort
covariates	an ffdf containing the covariates for each person in the cohort
covariateRef	an ffdf containing details about the covariates
metadata	a list containing details about the data extraction

See Also

DatabaseConnector, OhdsiRTools, SqlRender, ggplot2, reshape2, dplyr, plyr

Examples

# to extract the males ages between 30 and 45 in cdm_test.dbo.cohort with id 21
# and find whether they have the default concept definitions 1 to 60 days prior
# to cohort start:
dbconnection <- DatabaseConnector::createConnectionDetails(dbms = dbms,server = server,
user = user,password = pw,port = port,schema = cdmDatabaseSchema)

data <- dataExtract(dbconnection, cdmDatabaseSchema='cdm_test.dbo',
                    cohortDatabaseSchema='cdm_test.dbo', cohort_id=21,
                    minAge = 30, maxAge=45, gender=8507,
                    type='group', groupDef='default',
                    historyStart=1,historyEnd=60,
                    ffloc='C:fftemps')

# to extract the cluster data using user specified concept sets:
# where definition 1 contains concept_ids: 101,32011,1 and 63434
#       definition 2 contains concept_ids: 12,13
#       definition 3 contains concept_ids: 450453,21435324,232,3424,4534435 and 3453
groupDef <- data.frame(covariate=c(1,1,1,1,2,2,3,3,3,3,3,3),
                       concept_id =c(c(101,32011,1,63434), c(12,13),
                                     c(450453,21435324,232,3424,4534435,3453))
data <- dataExtract(dbconnection, cdmDatabaseSchema='cdm_test.dbo',
                    cohortDatabaseSchema='cdm_test.dbo', cohort_id=21,
                    type='group', groupDef=groupDef,
                    historyStart=1,historyEnd=180,
                    ffloc='C:fftemps')

jreps/patientCluster documentation built on May 20, 2019, 10:46 a.m.