CohortMethod v6.0.2 has been released, containing 2 minor bug fixes.
DataQualityDashboard v2.8.9 has been released. This release contains a bugfix for an issue which was causing multi-threaded runs to error out, along with some other minor fixes.
MethodEvaluation 2.4.1 has been released. This contains some very minor fixes.
PheValuator 2.2.17 has been released. It explicitly calls dplyr in CreateEvalCohort and corrects a sql error for sql server.
Andromeda 1.2.1 has been released. This fixes a bug throwing an unable-to-resize-memory error under some very specific conditions (which arise often when using PheValuator).
Keeper 2.1.3 has been released. This includes several bug fixes. Continuous integration now tests against all test database servers.
SelfControlledCohort v2.0.0 has been released. This is a major release (and a first CRAN release) that includes the addition of EmpiricalCalibration of effect estimates as well as study diagnostics to align with other HADES packages.
1 Like
CohortMethod 6.0.3 has been released. This fixes an error caused by the new dbplyr version 2.6.0.
FeatureExtraction v3.14.0 has been released. Next to a bug fix, this includes functionality to get features for observation.value_as_number. Furthermore, it is now possible to set a minCharacterizationCount to limit features (next to the already existing minCharacterizationMean)
DatabaseConnector 7.2.0 has been released. This contains driver updates for BigQuery, Iris, and Snowflake, and support for newer DataBricks drivers (see release notes) as well as bulk import from DataBricks.
This release also fixes all the dbplyr-related errors caused by dbplyr 2.6.0.
CohortGenerator v1.1.1 has been released. This contains a couple of bug fixes to better handle subset definitions using v0.x of the package and to prevent duplicate cohort ID collisions.
SqlRender 1.19.6 has been released. This includes minor bug fixes for DDLs in Spark (DataBricks).
CohortIncidence 4.2.0 has been released and added to CRAN. Full release information is in NEWS.md.
SqlRender 1.19.7 has been released. This contains some bugfixes for Iris, PostgreSQL, DuckDB and SQLite.
MethodEvaluation 2.4.2 has been released with some performance improvements for very large databases.
Capr v3 has been released. Lots of improvements including addition of all circe-be attributes, json to capr import (thanks @Adam_Black) and a new Capr agent mode thanks to @katy-sadowski!
5 Likes
That’s cool! Can you do a demo at the symposium workgroup meeting?
Keeper 2.2.0 has been released. Two changes:
- You can now provide a clinical definition of your phenotype when generating concept sets and reviewing patient profiles.
- Added a ‘hypernym’ category for a concept that is broader than the disease of interest. This is also used for constructing the highly-sensitive cohort when there is no data coded at the level of the disease of interest itself.
Strategus v1.6.0 has been released. This new release of Strategus includes a new module for the PheValuator package (thanks @Bill_OBrien!) and updates to support Characterization v4 and CohortIncidence v4.2 (thanks @jreps!). This release also contains additional features to report execution & upload status to resume work if a failure occurs.
PatientLevelPrediction v6.7.0 has been released. Highlights include Broken Adaptive Ridge logistic regression, semi-synthetic benchmark data generation, and training splits with a configurable number of outcomes. This release also includes fixes for iterative hard thresholding, cross-validation settings, and loading saved support vector machine models.
DeepPatientLevelPrediction v2.4.0 has also been released. This adds hyperparameter tuning for RealMLP and fixes a temporal data alignment issue when multiple features share the same timestamp. It also includes a new temporal transformer vignette and updated Python/PyTorch installation guidance.