Skip to content

Commit

Permalink
minor fixes
Browse files Browse the repository at this point in the history 
minor fixes
  • Loading branch information
@jreps
jreps committed Jul 19, 2024
1 parent 692b287 commit 0b61a14
Show file tree
Hide file tree
Showing 3 changed files with 64 additions and 39 deletions.
18 changes: 16 additions & 2 deletions R/AggregateCovariates.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -651,7 +651,7 @@ exportAndromedaToCsv <- function(
if (sum(removeInd) > 0) {
ParallelLogger::logInfo(paste0("Removing sum_value counts less than ", minCellCount))
if (sum(removeInd) > 0) {
data$sum_value[removeInd] <- -1
data$sum_value[removeInd] <- -1*minCellCount
}
}

Expand Down Expand Up @@ -683,7 +683,7 @@ exportAndromedaToCsv <- function(
if (sum(removeInd) > 0) {
ParallelLogger::logInfo(paste0("Removing count_value counts less than ", minCellCount))
if (sum(removeInd) > 0) {
data$count_value[removeInd] <- -1
data$count_value[removeInd] <- -1*minCellCount
}
}

Expand Down Expand Up @@ -725,6 +725,20 @@ exportAndromedaToCsv <- function(
colnames(cohortCounts) <- SqlRender::camelCaseToSnakeCase(colnames(cohortCounts))

# TODO apply minCellCount to columns row_count, person_count
removeInd <- cohortCounts$row_count < minCellCount
if (sum(removeInd) > 0) {
ParallelLogger::logInfo(paste0("Removing row_count counts less than ", minCellCount))
if (sum(removeInd) > 0) {
cohortCounts$row_count[removeInd] <- -1*minCellCount
}
}
removeInd <- cohortCounts$person_count < minCellCount
if (sum(removeInd) > 0) {
ParallelLogger::logInfo(paste0("Removing person_count counts less than ", minCellCount))
if (sum(removeInd) > 0) {
cohortCounts$person_count[removeInd] <- -1*minCellCount
}
}

if(file.exists(file.path(saveLocation, 'cohort_counts.csv'))){
append <- T
Expand Down
28 changes: 19 additions & 9 deletions R/RunCharacterization.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -439,17 +439,27 @@ aggregateCsvs <- function(

if(csvType == 'analysis_ref.csv'){
data <- data %>%
dplyr::filter( # need to filter analysis_id and covariate_setting_id
!.data$setting_id %in% analysisRefTracker
)
analysisRefTracker <- c(analysisRefTracker, unique(data$setting_id))
dplyr::mutate(
unique_id = paste0(.data$setting_id, '-', .data$analysis_id)
) %>%
dplyr::filter( # need to filter analysis_id and setting_id
!.data$unique_id %in% analysisRefTracker
) %>%
dplyr::select(-"unique_id")

analysisRefTracker <- unique(c(analysisRefTracker, paste0(data$setting_id,'-',data$analysis_id)))
}
if(csvType == 'covariate_ref.csv'){
if(csvType == 'covariate_ref.csv'){ # this could be problematic as may have differnet covariate_ids
data <- data %>%
dplyr::filter(
!.data$setting_id %in% covariateRefTracker
)
covariateRefTracker <- c(covariateRefTracker, unique(data$setting_id))
dplyr::mutate(
unique_id = paste0(.data$setting_id, '-', .data$covariate_id)
) %>%
dplyr::filter( # need to filter covariate_id and setting_id
!.data$unique_id %in% covariateRefTracker
)%>%
dplyr::select(-"unique_id")

covariateRefTracker <- unique(c(covariateRefTracker, paste0(data$setting_id,'-',data$covariate_id)))
}
if(csvType == 'settings.csv'){
data <- data %>%
Expand Down
57 changes: 29 additions & 28 deletions inst/sql/sql_server/DechallengeRechallenge.sql
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,25 +13,26 @@ select
@dechallenge_stop_interval as dechallenge_stop_interval,
@dechallenge_evaluation_window as dechallenge_evaluation_window,
target_cohort_definition_id,
outcome_cohort_definition_id,
num_exposure_eras,
num_persons_exposed,
num_cases,
dechallenge_attempt,
dechallenge_fail,
dechallenge_success,
rechallenge_attempt,
rechallenge_fail,
rechallenge_success,
case when num_cases > 0 then 1.0*dechallenge_attempt / num_cases else null end as pct_dechallenge_attempt,
case when dechallenge_attempt > 0 then 1.0*dechallenge_success / dechallenge_attempt else null end as pct_dechallenge_success,
case when dechallenge_attempt > 0 then 1.0*dechallenge_fail / dechallenge_attempt else null end as pct_dechallenge_fail,
case when dechallenge_attempt > 0 then 1.0*rechallenge_attempt / dechallenge_attempt else null end as pct_rechallenge_attempt,
case when rechallenge_attempt > 0 then 1.0*rechallenge_success / rechallenge_attempt else null end as pct_rechallenge_success,
case when rechallenge_attempt > 0 then 1.0*rechallenge_fail / rechallenge_attempt else null end as pct_rechallenge_fail
outcome_cohort_definition_id,
num_exposure_eras,
num_persons_exposed,
num_cases,
dechallenge_attempt,
dechallenge_fail,
dechallenge_success,
rechallenge_attempt,
rechallenge_fail,
rechallenge_success,
case when num_cases > 0 then 1.0*dechallenge_attempt / num_cases else null end as pct_dechallenge_attempt,
case when dechallenge_attempt > 0 then 1.0*dechallenge_success / dechallenge_attempt else null end as pct_dechallenge_success,
case when dechallenge_attempt > 0 then 1.0*dechallenge_fail / dechallenge_attempt else null end as pct_dechallenge_fail,
case when dechallenge_attempt > 0 then 1.0*rechallenge_attempt / dechallenge_attempt else null end as pct_rechallenge_attempt,
case when rechallenge_attempt > 0 then 1.0*rechallenge_success / rechallenge_attempt else null end as pct_rechallenge_success,
case when rechallenge_attempt > 0 then 1.0*rechallenge_fail / rechallenge_attempt else null end as pct_rechallenge_fail

into #challenge
from
INTO #challenge

FROM
(
select cases.target_cohort_definition_id, cases.outcome_cohort_definition_id,
exposures.num_exposure_eras,
Expand Down Expand Up @@ -69,15 +70,15 @@ from
and io1.cohort_start_date > dc1.cohort_start_date and io1.cohort_start_date <= dc1.cohort_end_date
group by dc1.cohort_definition_id, io1.cohort_definition_id

-- added this code to return 0s when there are no outcomes
-- so we can tell whether the dechal has been run or not
union

select distinct
dc1_temp.cohort_definition_id as target_cohort_definition_id,
io1_temp.cohort_definition_id as outcome_cohort_definition_id,
0 as num_cases
from #target_cohort dc1_temp
join #outcome_cohort io1_temp
cross join #outcome_cohort io1_temp

) temp_cases
group by
target_cohort_definition_id,
Expand Down Expand Up @@ -109,7 +110,7 @@ from
on dc1.subject_id = ro1.subject_id
and io1.cohort_definition_id = ro1.cohort_definition_id
and ro1.cohort_start_date > dc1.cohort_end_date
and ro1.cohort_start_date <= dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date) --this should be parameterized to be the dechallenge window required for success/failure
and ro1.cohort_start_date <= dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date)
group by dc1.cohort_definition_id, io1.cohort_definition_id
) dechallenge_fail
on cases.target_cohort_definition_id = dechallenge_fail.target_cohort_definition_id
Expand All @@ -126,12 +127,12 @@ from
on dc1.subject_id = ro0.subject_id
and io1.cohort_definition_id = ro0.cohort_definition_id
and ro0.cohort_start_date > dc1.cohort_end_date
and ro0.cohort_start_date <= dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date) --this should be parameterized to be the dechallenge window required for success/failure
and ro0.cohort_start_date <= dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date)
inner join #target_cohort de1
on dc1.subject_id = de1.subject_id
and dc1.cohort_definition_id = de1.cohort_definition_id
and de1.cohort_start_date > dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date) --using same dechallenge window to detrmine when rechallenge attempt can start
where ro0.subject_id is null --not a dechallenge fail
and de1.cohort_start_date > dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date)
where ro0.subject_id is null
group by dc1.cohort_definition_id, io1.cohort_definition_id
) rechallenge_attempt
on cases.target_cohort_definition_id = rechallenge_attempt.target_cohort_definition_id
Expand All @@ -148,17 +149,17 @@ from
on dc1.subject_id = ro0.subject_id
and io1.cohort_definition_id = ro0.cohort_definition_id
and ro0.cohort_start_date > dc1.cohort_end_date
and ro0.cohort_start_date <= dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date) --this should be parameterized to be the dechallenge window required for success/failure
and ro0.cohort_start_date <= dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date)
inner join #target_cohort de1
on dc1.subject_id = de1.subject_id
and dc1.cohort_definition_id = de1.cohort_definition_id
and de1.cohort_start_date > dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date) --using same dechallenge window to detrmine when rechallenge attempt can start
and de1.cohort_start_date > dateadd(day, @dechallenge_evaluation_window, dc1.cohort_end_date)
inner join #outcome_cohort ro1
on de1.subject_id = ro1.subject_id
and io1.cohort_definition_id = ro1.cohort_definition_id
and ro1.cohort_start_date > de1.cohort_start_date
and ro1.cohort_start_date <= de1.cohort_end_date
where ro0.subject_id is null --not a dechallenge fail
where ro0.subject_id is null
group by dc1.cohort_definition_id, io1.cohort_definition_id
) rechallenge_fail
on cases.target_cohort_definition_id = rechallenge_fail.target_cohort_definition_id
Expand Down

0 comments on commit 0b61a14

Please sign in to comment.