LegendT2dm 2.1.2
=======================
Changes:
- Small patch in
computeBalance
function for robustness
LegendT2dm 2.1.1
=======================
Changes:
1.Update covariate balance computation and results data model for covariate_balance
table
Changes:
- Enable drug-vs-drug comparisons for all drug classes (within and across classes)
- Fixed a bug for drug-level studies where OT2 studies failed to execute
- Include continuous age in patient characteristics
Changes:
- Handle ungraceful crash when all paired exposure cohorts are too small
Changes:
- Expand study package to include within-class drug-level CES studies for SGLT2 Inhibitors
Changes:
- Add meta-analysis routines
- Update
LegendT2dmEvidenceExplorer
to display meta-analysis results - Add
sources
tocohort_method_result
in results schema to record which sources are used in a meta-analysis
Changes:
- Add
database_id
as column to resultinglikelihood_profile
table
Changes:
- Change
Inf
numeric values toNA
for upload to database (TODO: Fix)
Changes:
- Fix
analysisSummary.csv
when using multiple runs
Changes:
- Add option to execute only
Run_1
Changes:
- Do not limit
Glycemic control
cohort to first occurrence - Add with-prior-outcome analyses for
Glycemic control
- Expose
endStudyDate
inexecute()
for EHR data sources with end-of-observation dates - Set
minDaysAtRisk
to 0 so that initial OT and ITT populations are the same to improve reproducibility as an ATLAS-generated study, since only a single PS model is constructed for both populations - Report # of persons with 0 days at risk in
cm_follow_up_table
- Include data/time of results in returned artifacts
- Fix negative control calibration in results export
Changes:
- Fix custom data pull
Changes:
- Patch
CohortDiagnostics
to exploit castcohortConceptIds
to 32-bit for BigQuery
Changes:
- Patch
CohortDiagnostics
to truncatedsum_value
andmean
columns incovariate_balance.csv
tominCellCount
- Add
NEWS.md
file - Patch path-specification error in
upload*
functions
Initial version