Feature Extraction

https://ohdsi.github.io/FeatureExtraction/
Admin
Feature Extraction is an R package for generating features (covariates) for a cohort using data in...
Tokens:46,278
Snippets:372
Trust Score:9.7
Update:4 months ago
Show doc for...
Tokens:
Raw
### Create Cohorts and Get Covariate Data

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDbCovariateData.html

This example demonstrates setting up connection details using Eunomia, creating cohorts, and then retrieving covariate data using getDbCovariateData. Ensure you have the Eunomia package installed and that the CDM database schema and cohort table are correctly specified.

```r
# \donttest{
eunomiaConnectionDetails <- Eunomia::getEunomiaConnectionDetails()
covSettings <- createDefaultCovariateSettings()
Eunomia::createCohorts(
  connectionDetails = eunomiaConnectionDetails,
  cdmDatabaseSchema = "main",
  cohortDatabaseSchema = "main",
  cohortTable = "cohort"
)
#> Cohorts created in table main.cohort
#>   cohortId       name
#> 1        1  Celecoxib
#> 2        2 Diclofenac
#> 3        3    GiBleed
#> 4        4     NSAIDs
#>                                                                                        description
#> 1    A simplified cohort definition for new users of celecoxib, designed specifically for Eunomia.
#> 2    A simplified cohort definition for new users ofdiclofenac, designed specifically for Eunomia.
#> 3 A simplified cohort definition for gastrointestinal bleeding, designed specifically for Eunomia.
#> 4       A simplified cohort definition for new users of NSAIDs, designed specifically for Eunomia.
#>   count
#> 1  1844
#> 2   850
#> 3   479
#> 4  2694
covData <- getDbCovariateData(
  connectionDetails = eunomiaConnectionDetails,
  tempEmulationSchema = NULL,
  cdmDatabaseSchema = "main",
  cdmVersion = "5",
  cohortTable = "cohort",
  cohortDatabaseSchema = "main",
  cohortTableIsTemp = FALSE,
  cohortIds = -1,
  rowIdField = "subject_id",
  covariateSettings = covSettings,
  aggregated = FALSE
)
#> Connecting using SQLite driver
#> Constructing features on server
#>   |                                                                              |                                                                      |   0%  |                                                                              |=                                                                     |   1%  |                                                                              |=                                                                     |   2%  |                                                                              |==                                                                    |   2%  |                                                                              |==                                                                    |   3%  |                                                                              |===                                                                   |   4%  |                                                                              |===                                                                   |   5%  |                                                                              |====                                                                  |   5%  |                                                                              |====                                                                  |   6%  |                                                                              |=====                                                                 |   7%  |                                                                              |=====                                                                 |   8%  |                                                                              |======
# }
```

--------------------------------

### Create Cohorts and Get Covariate Data

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDbCohortAttrCovariatesData.html

This example demonstrates setting up a connection, creating cohorts, defining covariate settings, and then extracting covariate data from the database using the getDbCohortAttrCovariatesData function. It requires the Eunomia and DatabaseConnector packages.

```r
# \donttest{
connectionDetails <- Eunomia::getEunomiaConnectionDetails()
Eunomia::createCohorts(
  connectionDetails = connectionDetails,
  cdmDatabaseSchema = "main",
  cohortDatabaseSchema = "main",
  cohortTable = "cohort"
)
#> Cohorts created in table main.cohort
#>   cohortId       name
#> 1        1  Celecoxib
#> 2        2 Diclofenac
#> 3        3    GiBleed
#> 4        4     NSAIDs
#>                                                                                        description
#> 1    A simplified cohort definition for new users of celecoxib, designed specifically for Eunomia.
#> 2    A simplified cohort definition for new users ofdiclofenac, designed specifically for Eunomia.
#> 3 A simplified cohort definition for gastrointestinal bleeding, designed specifically for Eunomia.
#> 4       A simplified cohort definition for new users of NSAIDs, designed specifically for Eunomia.
#>   count
#> 1  1844
#> 2   850
#> 3   479
#> 4  2694
connection <- DatabaseConnector::connect(connectionDetails)
#> Connecting using SQLite driver
covariateSettings <- createCohortAttrCovariateSettings(
  attrDatabaseSchema = "main",
  cohortAttrTable = "cohort_attribute",
  attrDefinitionTable = "attribute_definition",
  includeAttrIds = c(1),
  isBinary = FALSE,
  missingMeansZero = FALSE
)

covData <- getDbCohortAttrCovariatesData(
  connection = connection,
  tempEmulationSchema = NULL,
  cdmDatabaseSchema = "main",
  cdmVersion = "5",
  cohortTable = "cohort",
  cohortIds = 1,
  rowIdField = "subject_id",
  covariateSettings = covariateSettings,
  aggregated = FALSE
)
#> Constructing covariates from cohort attributes table
#> Inserting data took 0.0172 secs
#> Loading took 0.0838 secs
# }
```

--------------------------------

### R Example: Using createDefaultCovariateSettings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createDefaultCovariateSettings.html

An example demonstrating how to create covariate settings using createDefaultCovariateSettings with specified concept and covariate IDs.

```r
# \donttest{
covSettings <- createDefaultCovariateSettings(
  includedCovariateConceptIds = c(1),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(2),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c(1)
)
# }
```

--------------------------------

### Example: Generating Cohort Data and Table 1

Source: https://ohdsi.github.io/FeatureExtraction/reference/createTable1.html

This R code demonstrates a complete workflow: setting up Eunomia connection details, creating default covariate settings, generating cohorts, retrieving covariate data for a specific cohort, and preparing it for table generation. This example is useful for understanding the data preparation steps required before calling `createTable1`.

```r
# \donttest{
eunomiaConnectionDetails <- Eunomia::getEunomiaConnectionDetails()
covSettings <- createDefaultCovariateSettings()
Eunomia::createCohorts(
  connectionDetails = eunomiaConnectionDetails,
  cdmDatabaseSchema = "main",
  cohortDatabaseSchema = "main",
  cohortTable = "cohort"
)
#> Cohorts created in table main.cohort
#>   cohortId       name
#> 1        1  Celecoxib
#> 2        2 Diclofenac
#> 3        3    GiBleed
#> 4        4     NSAIDs
#>                                                                                        description
#> 1    A simplified cohort definition for new users of celecoxib, designed specifically for Eunomia.
#> 2    A simplified cohort definition for new users ofdiclofenac, designed specifically for Eunomia.
#> 3 A simplified cohort definition for gastrointestinal bleeding, designed specifically for Eunomia.
#> 4       A simplified cohort definition for new users of NSAIDs, designed specifically for Eunomia.
#>   count
#> 1  1844
#> 2   850
#> 3   479
#> 4  2694
covData1 <- getDbCovariateData(
  connectionDetails = eunomiaConnectionDetails,
  tempEmulationSchema = NULL,
  cdmDatabaseSchema = "main",
  cdmVersion = "5",
  cohortTable = "cohort",
  cohortDatabaseSchema = "main",
  cohortTableIsTemp = FALSE,
  cohortId = 1,
  rowIdField = "subject_id",
  covariateSettings = covSettings,
  aggregated = TRUE
)
#> Warning: cohortId argument has been deprecated, please use cohortIds
#> Connecting using SQLite driver
#> Constructing features on server
#>   |                                                                              |                                                                      |   0%  |                                                                              |                                                                      |   1%  |                                                                              |=                                                                     |   1%  |                                                                              |=                                                                     |   2%  |                                                                              |==                                                                    |   2%  |                                                                              |==                                                                    |   3%  |                                                                              |===                                                                   |   4%  |                                                                              |===                                                                   |   5%  |                                                                              |====                                                                  |   5%  |                                                                              |====                                                                  |   6%  |                                                                              |=====                                                   
```

--------------------------------

### Create Covariate Settings Example

Source: https://ohdsi.github.io/FeatureExtraction/reference/createCovariateSettings.html

This example demonstrates how to configure covariate settings using the createCovariateSettings function. It includes various demographic and condition-based covariates, specifying their inclusion and time windows.

```r
settings <- createCovariateSettings(
  useDemographicsGender = TRUE,
  useDemographicsAge = FALSE,
  useDemographicsAgeGroup = TRUE,
  useDemographicsRace = TRUE,
  useDemographicsEthnicity = TRUE,
  useDemographicsIndexYear = TRUE,
  useDemographicsIndexMonth = TRUE,
  useDemographicsPriorObservationTime = FALSE,
  useDemographicsPostObservationTime = FALSE,
  useDemographicsTimeInCohort = FALSE,
  useDemographicsIndexYearMonth = FALSE,
  useCareSiteId = FALSE,
  useConditionOccurrenceAnyTimePrior = FALSE,
  useConditionOccurrenceLongTerm = FALSE,
  useConditionOccurrenceMediumTerm = FALSE,
  useConditionOccurrenceShortTerm = FALSE,
  useConditionOccurrencePrimaryInpatientAnyTimePrior = FALSE,
  useConditionOccurrencePrimaryInpatientLongTerm = FALSE,
  useConditionOccurrencePrimaryInpatientMediumTerm = FALSE,
  useConditionOccurrencePrimaryInpatientShortTerm = FALSE,
  useConditionEraAnyTimePrior = FALSE,
  useConditionEraLongTerm = FALSE,
  useConditionEraMediumTerm = FALSE,
  useConditionEraShortTerm = FALSE,
  useConditionEraOverlapping = FALSE,
  useConditionEraStartLongTerm = FALSE,
  useConditionEraStartMediumTerm = FALSE,
  useConditionEraStartShortTerm = FALSE,
  useConditionGroupEraAnyTimePrior = FALSE,
  useConditionGroupEraLongTerm = TRUE,
  useConditionGroupEraMediumTerm = FALSE,
  useConditionGroupEraShortTerm = TRUE,
  useConditionGroupEraOverlapping = FALSE,
  useConditionGroupEraStartLongTerm = FALSE,
  useConditionGroupEraStartMediumTerm = FALSE,
  useConditionGroupEraStartShortTerm = FALSE,
  useDrugExposureAnyTimePrior = FALSE,
  useDrugExposureLongTerm = FALSE,
  useDrugExposureMediumTerm = FALSE,
  useDrugExposureShortTerm = FALSE,
  useDrugEraAnyTimePrior = FALSE,
  useDrugEraLongTerm = FALSE,
  useDrugEraMediumTerm = FALSE,
  useDrugEraShortTerm = FALSE,
  useDrugEraOverlapping = FALSE,
  useDrugEraStartLongTerm = FALSE,
  useDrugEraStartMediumTerm = FALSE,
  useDrugEraStartShortTerm = FALSE,
  useDrugGroupEraAnyTimePrior = FALSE,
  useDrugGroupEraLongTerm = TRUE,
  useDrugGroupEraMediumTerm = FALSE,
  useDrugGroupEraShortTerm = TRUE,
  useDrugGroupEraOverlapping = TRUE,
  useDrugGroupEraStartLongTerm = FALSE,
)
```

--------------------------------

### Example Usage of createTable1CovariateSettings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createTable1CovariateSettings.html

Demonstrates how to use the createTable1CovariateSettings function to generate covariate settings for a Table 1. This example shows setting various parameters to their default or empty values.

```r
# \donttest{
table1CovSettings <- createTable1CovariateSettings(
  specifications = getDefaultTable1Specifications(),
  covariateSettings = createDefaultCovariateSettings(),
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)
# }
```

--------------------------------

### Example Usage of createDetailedTemporalCovariateSettings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createDetailedTemporalCovariateSettings.html

This example demonstrates how to use `createDetailedTemporalCovariateSettings` by first creating analysis details and then passing them to the function along with the desired temporal window.

```r
# \donttest{
analysisDetails <- createAnalysisDetails(
  analysisId = 1,
  sqlFileName = "DemographicsGender.sql",
  parameters = list(
    analysisId = 1,
    analysisName = "Gender",
    domainId = "Demographics"
  ),
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)
covSettings <- createDetailedTemporalCovariateSettings(
  analyses = analysisDetails,
  temporalStartDays = -365:-1,
  temporalEndDays = -365:-1
)
# }
```

--------------------------------

### Example Usage of createAnalysisDetails

Source: https://ohdsi.github.io/FeatureExtraction/reference/createAnalysisDetails.html

This example demonstrates how to use the createAnalysisDetails function to create settings for a 'Gender' covariate. It specifies the SQL file, parameters, and concept IDs to include or exclude.

```r
analysisDetails <- createAnalysisDetails(
  analysisId = 1,
  sqlFileName = "DemographicsGender.sql",
  parameters = list(
    analysisId = 1,
    analysisName = "Gender",
    domainId = "Demographics"
  ),
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)
```

--------------------------------

### Example Usage of getDefaultTable1Specifications

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDefaultTable1Specifications.html

Demonstrates how to obtain the default table 1 specifications and store them in a variable for later use with the `createTable1` function.

```r
# \donttest{
defaultTable1Specs <- getDefaultTable1Specifications()
# }
```

--------------------------------

### Example Covariate Settings Function

Source: https://ohdsi.github.io/FeatureExtraction/articles/CreatingCustomCovariateBuilders.html

An example of a function that creates a covariateSettings object. This object specifies the parameters for covariate generation and the name of the function that will perform the actual data extraction.

```r
createLooCovariateSettings <- function(useLengthOfObs = TRUE) {
  covariateSettings <- list(useLengthOfObs = useLengthOfObs)
  attr(covariateSettings, "fun") <- "getDbLooCovariateData"
  class(covariateSettings) <- "covariateSettings"
  return(covariateSettings)
}
```

--------------------------------

### Example Usage of computeStandardizedDifference

Source: https://ohdsi.github.io/FeatureExtraction/reference/computeStandardizedDifference.html

This example demonstrates how to load covariate data and then compute the standardized difference between two cohorts using `computeStandardizedDifference`. Ensure covariate data is loaded before calling the function.

```R
# \donttest{
binaryCovDataFile <- system.file("testdata/binaryCovariateData.zip",
  package = "FeatureExtraction"
)
covariateData1 <- loadCovariateData(binaryCovDataFile)
covariateData2 <- loadCovariateData(binaryCovDataFile)
covDataDiff <- computeStandardizedDifference(
  covariateData1,
  covariateData2,
  cohortId1 = 1,
  cohortId2 = 2
)
# }
```

--------------------------------

### Install FeatureExtraction Package

Source: https://ohdsi.github.io/FeatureExtraction/index.html

Install the FeatureExtraction package and its dependencies from the OHDSI repository. Ensure you have RTools and Java installed.

```r
install.packages("drat")
drat::addRepo("OHDSI")
install.packages("FeatureExtraction")
```

--------------------------------

### Database Connection Details (PostgreSQL)

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Configures connection details for a PostgreSQL database. This setup is required for the CohortMethod package to access the data.

```R
connectionDetails <- createConnectionDetails(
  dbms = "postgresql",
  server = "localhost/ohdsi",
  user = "joe",
  password = "supersecret"
)
```

--------------------------------

### Create Default Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Generates a comprehensive set of default covariates. This is the simplest way to start.

```r
settings <- createDefaultCovariateSettings()
```

--------------------------------

### R Example for createCohortAttrCovariateSettings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createCohortAttrCovariateSettings.html

Demonstrates how to use createCohortAttrCovariateSettings to define covariate settings based on cohort attributes. Ensure the specified database schema and tables exist.

```r
# \donttest{
covariateSettings <- createCohortAttrCovariateSettings(
  analysisId = 1,
  attrDatabaseSchema = "main",
  attrDefinitionTable = "attribute_definition",
  cohortAttrTable = "cohort_attribute",
  includeAttrIds = c(1),
  isBinary = FALSE,
  missingMeansZero = FALSE
)
# }
```

--------------------------------

### Example Usage of createDetailedCovariateSettings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createDetailedCovariateSettings.html

Demonstrates how to create analysis details and then use them to generate detailed covariate settings. This is intended for advanced users.

```r
# \donttest{
analysisDetails <- createAnalysisDetails(
  analysisId = 1,
  sqlFileName = "DemographicsGender.sql",
  parameters = list(
    analysisId = 1,
    analysisName = "Gender",
    domainId = "Demographics"
  ),
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)
covSettings <- createDetailedCovariateSettings(analyses = analysisDetails)
# }
```

--------------------------------

### R Example: Creating Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createDefaultTemporalCovariateSettings.html

Demonstrates how to use createDefaultTemporalCovariateSettings to define covariate settings, specifying included and excluded concept IDs and covariate IDs.

```r
# \donttest{
covSettings <- createDefaultTemporalCovariateSettings(
  includedCovariateConceptIds = c(1),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(2),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c(1)
)
# }
```

--------------------------------

### Example Usage of tidyCovariateData

Source: https://ohdsi.github.io/FeatureExtraction/reference/tidyCovariateData.html

Demonstrates how to use the tidyCovariateData function to clean and normalize covariate data. Requires creating an empty covariate data object first.

```r
# \donttest{
covariateData <- FeatureExtraction::createEmptyCovariateData(
  cohortIds = 1,
  aggregated = FALSE,
  temporal = FALSE
)

covData <- tidyCovariateData(
  covariateData = covariateData,
  minFraction = 0.001,
  normalize = TRUE,
  removeRedundancy = TRUE
)
#> Tidying covariates took 1.71 secs
# }
```

--------------------------------

### Default Temporal Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Creates default settings for temporal covariates. This is a convenient way to start when using temporal covariates.

```R
settings <- createDefaultTemporalCovariateSettings()
```

--------------------------------

### Create Temporal Sequence Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createTemporalSequenceCovariateSettings.html

Generates covariate settings for temporal sequence analysis. This example includes demographic features, procedure, device, measurement, and observation data, with a time window of -730 to -1 days relative to the index date.

```R
settings <- createTemporalSequenceCovariateSettings(
  useDemographicsGender = TRUE,
  useDemographicsAge = FALSE,
  useDemographicsAgeGroup = TRUE,
  useDemographicsRace = TRUE,
  useDemographicsEthnicity = TRUE,
  useDemographicsIndexYear = TRUE,
  useDemographicsIndexMonth = TRUE,
  useConditionOccurrence = FALSE,
  useConditionOccurrencePrimaryInpatient = FALSE,
  useConditionEraStart = FALSE,
  useConditionEraGroupStart = FALSE,
  useDrugExposure = FALSE,
  useDrugEraStart = FALSE,
  useDrugEraGroupStart = FALSE,
  useProcedureOccurrence = TRUE,
  useDeviceExposure = TRUE,
  useMeasurement = TRUE,
  useMeasurementValue = FALSE,
  useObservation = TRUE,
  timePart = "DAY",
  timeInterval = 1,
  sequenceEndDay = -1,
  sequenceStartDay = -730,
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)

```

--------------------------------

### Load Covariate Data from File

Source: https://ohdsi.github.io/FeatureExtraction/reference/loadCovariateData.html

Loads an object of class CovariateData from a zip file. Ensure the FeatureExtraction package is installed and loaded.

```r
binaryCovDataFile <- system.file("testdata/binaryCovariateData.zip",
  package = "FeatureExtraction"
)
covData <- loadCovariateData(binaryCovDataFile)
```

--------------------------------

### SQL for Creating Cohort Attributes and Definitions

Source: https://ohdsi.github.io/FeatureExtraction/articles/CreatingCovariatesUsingCohortAttributes.html

This SQL script creates a cohort attribute table and an attribute definition table. It calculates the length of observation prior to the cohort start date.

```sql
-----------------------------------
File LengthOfObsCohortAttr.sql 
-----------------------------------
IF OBJECT_ID('@cohort_database_schema.@cohort_attribute_table', 'U') IS NOT NULL
    DROP TABLE @cohort_database_schema.@cohort_attribute_table;
    
IF OBJECT_ID('@cohort_database_schema.@attribute_definition_table', 'U') IS NOT NULL
    DROP TABLE @cohort_database_schema.@attribute_definition_table;
    
SELECT cohort_definition_id,
    subject_id,
    cohort_start_date,
    1 AS attribute_definition_id,
    DATEDIFF(DAY, observation_period_start_date, cohort_start_date) AS value_as_number
INTO @cohort_database_schema.@cohort_attribute_table
FROM @cohort_database_schema.@cohort_table cohort
INNER JOIN @cdm_database_schema.observation_period op
    ON op.person_id = cohort.subject_id
WHERE cohort_start_date >= observation_period_start_date
    AND cohort_start_date <= observation_period_end_date
{@cohort_definition_ids != ''} ? {
    AND cohort_definition_id IN (@cohort_definition_ids)
}
;
    
SELECT 1 AS attribute_definition_id,
  'Length of observation in days' AS attribute_name
INTO @cohort_database_schema.@attribute_definition_table;
```

--------------------------------

### Create Cohorts of Interest SQL

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

SQL script to create a table of cohorts of interest based on drug eras and observation periods. It requires 365 days of observation prior to the cohort start date.

```sql
/***********************************
File cohortsOfInterest.sql
**********************************/

IF OBJECT_ID('@resultsDatabaseSchema.cohorts_of_interest', 'U') IS NOT NULL
DROP TABLE @resultsDatabaseSchema.cohorts_of_interest;

SELECT first_use.*
INTO @resultsDatabaseSchema.cohorts_of_interest
FROM (
SELECT drug_concept_id AS cohort_definition_id,
MIN(drug_era_start_date) AS cohort_start_date,
MIN(drug_era_end_date) AS cohort_end_date,
person_id 
FROM @cdmDatabaseSchema.drug_era
WHERE drug_concept_id = 1118084-- celecoxib
OR drug_concept_id = 1124300 --diclofenac
GROUP BY drug_concept_id, 
person_id
) first_use 
INNER JOIN @cdmDatabaseSchema.observation_period
ON first_use.person_id = observation_period.person_id
AND cohort_start_date >= observation_period_start_date
AND cohort_end_date <= observation_period_end_date
WHERE DATEDIFF(DAY, observation_period_start_date, cohort_start_date) >= 365;
```

--------------------------------

### Default Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createCovariateSettings.html

This snippet shows the default settings for creating covariates. It includes configurations for drug eras, procedure occurrences, device exposures, measurements, observations, and various comorbidity indices. Use this as a starting point for defining your covariate generation strategy.

```R
createCovariateSettings(
  useDrugGroupEraStartMediumTerm = FALSE,
  useDrugGroupEraStartShortTerm = FALSE,
  useProcedureOccurrenceAnyTimePrior = FALSE,
  useProcedureOccurrenceLongTerm = TRUE,
  useProcedureOccurrenceMediumTerm = FALSE,
  useProcedureOccurrenceShortTerm = TRUE,
  useDeviceExposureAnyTimePrior = FALSE,
  useDeviceExposureLongTerm = TRUE,
  useDeviceExposureMediumTerm = FALSE,
  useDeviceExposureShortTerm = TRUE,
  useMeasurementAnyTimePrior = FALSE,
  useMeasurementLongTerm = TRUE,
  useMeasurementMediumTerm = FALSE,
  useMeasurementShortTerm = TRUE,
  useMeasurementValueAnyTimePrior = FALSE,
  useMeasurementValueLongTerm = FALSE,
  useMeasurementValueMediumTerm = FALSE,
  useMeasurementValueShortTerm = FALSE,
  useMeasurementRangeGroupAnyTimePrior = FALSE,
  useMeasurementRangeGroupLongTerm = TRUE,
  useMeasurementRangeGroupMediumTerm = FALSE,
  useMeasurementRangeGroupShortTerm = TRUE,
  useMeasurementValueAsConceptAnyTimePrior = FALSE,
  useMeasurementValueAsConceptLongTerm = TRUE,
  useMeasurementValueAsConceptMediumTerm = FALSE,
  useMeasurementValueAsConceptShortTerm = TRUE,
  useObservationAnyTimePrior = FALSE,
  useObservationLongTerm = TRUE,
  useObservationMediumTerm = FALSE,
  useObservationShortTerm = TRUE,
  useObservationValueAsConceptAnyTimePrior = FALSE,
  useObservationValueAsConceptLongTerm = TRUE,
  useObservationValueAsConceptMediumTerm = FALSE,
  useObservationValueAsConceptShortTerm = TRUE,
  useCharlsonIndex = TRUE,
  useDcsi = TRUE,
  useChads2 = TRUE,
  useChads2Vasc = TRUE,
  useHfrs = FALSE,
  useDistinctConditionCountLongTerm = FALSE,
  useDistinctConditionCountMediumTerm = FALSE,
  useDistinctConditionCountShortTerm = FALSE,
  useDistinctIngredientCountLongTerm = FALSE,
  useDistinctIngredientCountMediumTerm = FALSE,
  useDistinctIngredientCountShortTerm = FALSE,
  useDistinctProcedureCountLongTerm = FALSE,
  useDistinctProcedureCountMediumTerm = FALSE,
  useDistinctProcedureCountShortTerm = FALSE,
  useDistinctMeasurementCountLongTerm = FALSE,
  useDistinctMeasurementCountMediumTerm = FALSE,
  useDistinctMeasurementCountShortTerm = FALSE,
  useDistinctObservationCountLongTerm = FALSE,
  useDistinctObservationCountMediumTerm = FALSE,
  useDistinctObservationCountShortTerm = FALSE,
  useVisitCountLongTerm = FALSE,
  useVisitCountMediumTerm = FALSE,
  useVisitCountShortTerm = FALSE,
  useVisitConceptCountLongTerm = FALSE,
  useVisitConceptCountMediumTerm = FALSE,
  useVisitConceptCountShortTerm = FALSE,
  longTermStartDays = -365,
  mediumTermStartDays = -180,
  shortTermStartDays = -30,
  endDays = 0,
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)
```

--------------------------------

### Setting up Eunomia and Creating Cohorts

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDbDefaultCovariateData.html

This snippet demonstrates how to establish a connection to the Eunomia database and create cohort tables. It is a prerequisite for running covariate extraction functions.

```r
# \donttest{
connectionDetails <- Eunomia::getEunomiaConnectionDetails()
Eunomia::createCohorts(
  connectionDetails = connectionDetails,
  cdmDatabaseSchema = "main",
  cohortDatabaseSchema = "main",
  cohortTable = "cohort"
)
#> Cohorts created in table main.cohort
#>   cohortId       name
#> 1        1  Celecoxib
#> 2        2 Diclofenac
#> 3        3    GiBleed
#> 4        4     NSAIDs
#>                                                                                        description
#> 1    A simplified cohort definition for new users of celecoxib, designed specifically for Eunomia.
#> 2    A simplified cohort definition for new users ofdiclofenac, designed specifically for Eunomia.
#> 3 A simplified cohort definition for gastrointestinal bleeding, designed specifically for Eunomia.
#> 4       A simplified cohort definition for new users of NSAIDs, designed specifically for Eunomia.
#>   count
#> 1  1844
#> 2   850
#> 3   479
#> 4  2694

```

--------------------------------

### Executing SQL and Fetching Data

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDbCovariateData.html

Shows the process of executing SQL queries and fetching data from the server, along with the time taken for each step.

```text
#> Executing SQL took 2.06 secs
#> Fetching data from server
#> Fetching data took 0.671 secs
# }
```

--------------------------------

### Get Deleted Redundant Covariate IDs

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Retrieves the IDs of covariates that were removed because they were redundant.

```r
deletedCovariateIds <- attr(tidyCovariates, "metaData")$deletedRedundantCovariateIds
```

--------------------------------

### getDbCovariateData

Source: https://ohdsi.github.io/FeatureExtraction/reference/index.html

Get covariate information from the database. This is a general function to retrieve covariate data from the database.

```APIDOC
## getDbCovariateData

### Description
Get covariate information from the database.

### Method
Not applicable (function call)

### Endpoint
Not applicable (function call)

### Parameters
None explicitly documented.

### Request Example
```
getDbCovariateData()
```

### Response
Not explicitly documented. Likely returns covariate data from the database.
```

--------------------------------

### Database Query Execution and Data Fetching

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDbDefaultCovariateData.html

Shows the process of executing a SQL query against a database and fetching the resulting data. Includes timing information for query execution and data retrieval.

```r
#> Executing SQL took 2.06 secs
#> Fetching data from server
#> Fetching data took 0.635 secs
# }
```

--------------------------------

### getDbDefaultCovariateData

Source: https://ohdsi.github.io/FeatureExtraction/reference/index.html

Get default covariate information from the database. This function retrieves default covariate data stored in the database.

```APIDOC
## getDbDefaultCovariateData

### Description
Get default covariate information from the database.

### Method
Not applicable (function call)

### Endpoint
Not applicable (function call)

### Parameters
None explicitly documented.

### Request Example
```
getDbDefaultCovariateData()
```

### Response
Not explicitly documented. Likely returns default covariate data from the database.
```

--------------------------------

### Get Default Table 1 Specifications

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDefaultTable1Specifications.html

Call this function to retrieve the default specifications object for table 1 creation.

```r
getDefaultTable1Specifications()
```

--------------------------------

### Execute Parameterized SQL with SqlRender

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

R code to read, render, translate, and execute parameterized SQL. This allows for schema and dialect flexibility.

```r
library(SqlRender)
sql <- readSql("cohortsOfInterest.sql")
sql <- render(sql,
  cdmDatabaseSchema = cdmDatabaseSchema,
  resultsDatabaseSchema = resultsDatabaseSchema
)
sql <- translate(sql, targetDialect = connectionDetails$dbms)

connection <- connect(connectionDetails)
executeSql(connection, sql)
```

--------------------------------

### getDbCohortAttrCovariatesData

Source: https://ohdsi.github.io/FeatureExtraction/reference/index.html

Get covariate information from the database through the cohort_attribute table. This function retrieves specific covariate data related to cohort attributes.

```APIDOC
## getDbCohortAttrCovariatesData

### Description
Get covariate information from the database through the cohort_attribute table.

### Method
Not applicable (function call)

### Endpoint
Not applicable (function call)

### Parameters
None explicitly documented.

### Request Example
```
getDbCohortAttrCovariatesData()
```

### Response
Not explicitly documented. Likely returns covariate data from the cohort_attribute table.
```

--------------------------------

### Get Deleted Infrequent Covariate IDs

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Retrieves the IDs of covariates that were removed due to being infrequent (appearing in less than minFraction of subjects).

```r
deletedCovariateIds <- attr(tidyCovariates, "metaData")$deletedInfrequentCovariateIds
```

--------------------------------

### Generate Table 1 with Custom Specifications

Source: https://ohdsi.github.io/FeatureExtraction/reference/createTable1.html

Use this snippet to create a Table 1 by providing two covariate datasets, cohort IDs, and custom specifications. It demonstrates how to control output formatting like counts, percentages, and decimal places.

```r
table1 <- createTable1(
  covariateData1 = covData1,
  covariateData2 = covData2,
  cohortId1 = 1,
  cohortId2 = 2,
  specifications = getDefaultTable1Specifications(),
  output = "one column",
  showCounts = FALSE,
  showPercent = TRUE,
  percentDigits = 1,
  valueDigits = 1,
  stdDiffDigits = 2
)
```

--------------------------------

### Create Default Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Generates default settings for creating covariates. This is the first step before fetching covariate data.

```r
covariateSettings <- createDefaultCovariateSettings()
```

--------------------------------

### Execute SQL with R

Source: https://ohdsi.github.io/FeatureExtraction/articles/CreatingCovariatesBasedOnOtherCohorts.html

This R code reads an SQL file, connects to the database, and executes the SQL script using SqlRender for translation and execution.

```r
library(SqlRender)
sql <- readSql("covariateCohorts.sql")
connection <- connect(connectionDetails)
renderTranslateExecuteSql(
  connection = connection,
  sql = sql,
  cdm_database_schema = cdmDatabaseSchema,
  cohort_database_schema = cohortDatabaseSchema,
  cohort_table = cohortTable
)
```

--------------------------------

### Create Detailed Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createDetailedCovariateSettings.html

Basic function signature for creating detailed covariate settings.

```r
createDetailedCovariateSettings(analyses = list())
```

--------------------------------

### Get Covariate Data for Table 1

Source: https://ohdsi.github.io/FeatureExtraction/reference/createTable1.html

Retrieves covariate data from the database to be used for generating Table 1. Ensure connection details and cohort settings are properly configured.

```r
covData2 <- getDbCovariateData(
  connectionDetails = eunomiaConnectionDetails,
  tempEmulationSchema = NULL,
  cdmDatabaseSchema = "main",
  cdmVersion = "5",
  cohortTable = "cohort",
  cohortDatabaseSchema = "main",
  cohortTableIsTemp = FALSE,
  cohortId = 2,
  rowIdField = "subject_id",
  covariateSettings = covSettings,
  aggregated = TRUE
)
```

--------------------------------

### Define and Populate Cohort Tables

Source: https://ohdsi.github.io/FeatureExtraction/articles/CreatingCovariatesBasedOnOtherCohorts.html

This SQL script defines two cohorts: one for diclofenac initiation and another for type 2 diabetes. It creates and inserts data into a specified cohort table.

```sql
/************************
File covariateCohorts.sql
*************************/

DROP TABLE IF EXISTS @cohort_database_schema.@cohort_table;

CREATE TABLE @cohort_database_schema.@cohort_table (
    cohort_definition_id INT,
    subject_id BIGINT,
    cohort_start_date DATE,
    cohort_end_date DATE
    );

INSERT INTO @cohort_database_schema.@cohort_table (
    cohort_definition_id,
    subject_id,
    cohort_start_date,
    cohort_end_date
    )
SELECT 1,
    person_id,
    MIN(drug_era_start_date),
    MIN(drug_era_end_date)
FROM @cdm_database_schema.drug_era
WHERE drug_concept_id = 1124300 --diclofenac
GROUP BY person_id;

INSERT INTO @cohort_database_schema.@cohort_table (
    cohort_definition_id,
    subject_id,
    cohort_start_date,
    cohort_end_date
    )
SELECT 2,
    condition_occurrence.person_id,
    MIN(condition_start_date),
    MIN(observation_period_end_date)
FROM @cdm_database_schema.condition_occurrence
INNER JOIN @cdm_database_schema.drug_exposure
    ON condition_occurrence.person_id = drug_exposure.person_id
        AND drug_exposure_start_date >= condition_start_date
        AND drug_exposure_start_date < DATEADD(DAY, 30, condition_start_date)
INNER JOIN @cdm_database_schema.observation_period
    ON condition_occurrence.person_id = observation_period.person_id
        AND condition_start_date >= observation_period_start_date
        AND condition_start_date <= observation_period_end_date
WHERE condition_concept_id IN (
        SELECT descendant_concept_id
        FROM @cdm_database_schema.concept_ancestor
        WHERE ancestor_concept_id = 201826 -- Type 2 diabetes mellitus
        )
    AND drug_concept_id IN (
        SELECT descendant_concept_id
        FROM @cdm_database_schema.concept_ancestor
        WHERE ancestor_concept_id = 21600712 -- DRUGS USED IN DIABETES (ATC A10)
        )
GROUP BY condition_occurrence.person_id;
```

--------------------------------

### Get Covariate Data for Cohort

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Extract covariate data from the CDM database for a specified cohort and covariate settings. Ensure connection details, schemas, and cohort information are correctly provided.

```r
covCelecoxib <- getDbCovariateData(
  connectionDetails = connectionDetails,
  cdmDatabaseSchema = cdmDatabaseSchema,
  cohortDatabaseSchema = resultsDatabaseSchema,
  cohortTable = "cohorts_of_interest",
  cohortIds = c(1118084),
  covariateSettings = settings,
  aggregated = TRUE
)
```

```r
covDiclofenac <- getDbCovariateData(
  connectionDetails = connectionDetails,
  cdmDatabaseSchema = cdmDatabaseSchema,
  cohortDatabaseSchema = resultsDatabaseSchema,
  cohortTable = "cohorts_of_interest",
  cohortIds = c(1118084),
  covariateSettings = settings,
  aggregated = TRUE
)
```

--------------------------------

### Rendering and Executing SQL with SqlRender

Source: https://ohdsi.github.io/FeatureExtraction/articles/CreatingCovariatesUsingCohortAttributes.html

This R code uses SqlRender to read, render, and translate SQL, then executes it against a database connection. Ensure connection details and schemas are correctly configured.

```r
library(SqlRender)
sql <- readSql("LengthOfObsCohortAttr.sql")
sql <- render(sql,
  cdm_database_schema = cdmDatabaseSchema,
  cohort_database_schema = cohortDatabaseSchema,
  cohort_table = "rehospitalization",
  cohort_attribute_table = "loo_cohort_attr",
  attribute_definition_table = "loo_attr_def",
  cohort_definition_ids = c(1, 2)
)
sql <- translate(sql, targetDialect = connectionDetails$dbms)

connection <- connect(connectionDetails)
executeSql(connection, sql)
```

--------------------------------

### getDefaultTable1Specifications

Source: https://ohdsi.github.io/FeatureExtraction/reference/getDefaultTable1Specifications.html

Loads the default specifications for a table 1, to be used with the createTable1 function.

```APIDOC
## getDefaultTable1Specifications

### Description
Loads the default specifications for a table 1, to be used with the `createTable1` function.

### Method
```R
getDefaultTable1Specifications()
```

### Returns
A specifications object.
```

--------------------------------

### Prespecified Temporal Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Creates temporal covariate settings using prespecified options for condition occurrence and measurement values. By default, covariates are created daily for the 365 days prior to cohort start.

```R
settings <- createTemporalCovariateSettings(
  useConditionOccurrence = TRUE,
  useMeasurementValue = TRUE
)
```

--------------------------------

### Get Per-Person Covariate Data

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Fetches per-person covariate data for a specified cohort. Requires connection details, CDM schema, cohort schema, cohort table, cohort IDs, row ID field, and covariate settings.

```r
covariateData <- getDbCovariateData(
  connectionDetails = connectionDetails,
  cdmDatabaseSchema = cdmDatabaseSchema,
  cohortDatabaseSchema = resultsDatabaseSchema,
  cohortTable = "cohorts_of_interest",
  cohortIds = c(1118084),
  rowIdField = "subject_id",
  covariateSettings = covariateSettings
)
```

--------------------------------

### createAnalysisDetails

Source: https://ohdsi.github.io/FeatureExtraction/reference/index.html

Create detailed covariate settings. This function is used to define detailed settings for creating covariates.

```APIDOC
## createAnalysisDetails

### Description
Create detailed covariate settings.

### Method
Not applicable (function call)

### Endpoint
Not applicable (function call)

### Parameters
None explicitly documented.

### Request Example
```
createAnalysisDetails()
```

### Response
Not explicitly documented. Likely returns detailed covariate settings.
```

--------------------------------

### Create Temporal Covariate Settings

Source: https://ohdsi.github.io/FeatureExtraction/reference/createTemporalCovariateSettings.html

This snippet demonstrates how to configure various demographic and clinical event covariates for temporal analysis. It specifies which types of data to include and defines the temporal windows for covariate generation.

```r
settings <- createTemporalCovariateSettings(
  useDemographicsGender = TRUE,
  useDemographicsAge = FALSE,
  useDemographicsAgeGroup = TRUE,
  useDemographicsRace = TRUE,
  useDemographicsEthnicity = TRUE,
  useDemographicsIndexYear = TRUE,
  useDemographicsIndexMonth = TRUE,
  useDemographicsPriorObservationTime = FALSE,
  useDemographicsPostObservationTime = FALSE,
  useDemographicsTimeInCohort = FALSE,
  useDemographicsIndexYearMonth = FALSE,
  useCareSiteId = FALSE,
  useConditionOccurrence = FALSE,
  useConditionOccurrencePrimaryInpatient = FALSE,
  useConditionEraStart = FALSE,
  useConditionEraOverlap = FALSE,
  useConditionEraGroupStart = FALSE,
  useConditionEraGroupOverlap = TRUE,
  useDrugExposure = FALSE,
  useDrugEraStart = FALSE,
  useDrugEraOverlap = FALSE,
  useDrugEraGroupStart = FALSE,
  useDrugEraGroupOverlap = TRUE,
  useProcedureOccurrence = TRUE,
  useDeviceExposure = TRUE,
  useMeasurement = TRUE,
  useMeasurementValue = FALSE,
  useMeasurementRangeGroup = TRUE,
  useMeasurementValueAsConcept = TRUE,
  useObservation = TRUE,
  useObservationValueAsConcept = TRUE,
  useCharlsonIndex = TRUE,
  useDcsi = TRUE,
  useChads2 = TRUE,
  useChads2Vasc = TRUE,
  useHfrs = FALSE,
  useDistinctConditionCount = FALSE,
  useDistinctIngredientCount = FALSE,
  useDistinctProcedureCount = FALSE,
  useDistinctMeasurementCount = FALSE,
  useDistinctObservationCount = FALSE,
  useVisitCount = FALSE,
  useVisitConceptCount = FALSE,
  temporalStartDays = -365:-1,
  temporalEndDays = -365:-1,
  includedCovariateConceptIds = c(),
  addDescendantsToInclude = FALSE,
  excludedCovariateConceptIds = c(),
  addDescendantsToExclude = FALSE,
  includedCovariateIds = c()
)
```

--------------------------------

### Custom Temporal Covariate Settings with Intervals

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Creates temporal covariate settings with custom time windows defined as 7-day intervals. This allows for a more granular temporal resolution.

```R
settings <- createTemporalCovariateSettings(
  useConditionOccurrence = TRUE,
  useMeasurementValue = TRUE,
  temporalStartDays = seq(-364, -7, by = 7),
  temporalEndDays = seq(-358, -1, by = 7)
)
```

--------------------------------

### Create Table 1 Comparison

Source: https://ohdsi.github.io/FeatureExtraction/articles/UsingFeatureExtraction.html

Generate a standard Table 1 comparison between two sets of covariate data, outputting results in two columns for easy side-by-side viewing.

```r
result <- createTable1(
  covariateData1 = covCelecoxib,
  covariateData2 = covDiclofenac,
  output = "two columns"
)
print(result, row.names = FALSE, right = FALSE)
```
Tokens:
Raw