Releases: knatarajan-lab/cehrbert_data
Releases · knatarajan-lab/cehrbert_data
v0.0.5
What's Changed
- Ehrshot conversion by @ChaoPang in #11
- added inpatient_att_type as an option to generate_training_data as well as create spark cohorts by @ChaoPang in #12
- Revert "added inpatient_att_type as an option to generate_training_data as well as create spark cohorts" by @ChaoPang in #13
- added inpatient_att_type as an option to generate_training_data as well as create spark cohorts by @ChaoPang in #14
- Ehrshot conversion enhancement by @ChaoPang in #15
- Bound visit end date by @ChaoPang in #16
- fixed a window function in defining the cohort sample partition by @ChaoPang in #17
- Fixed the union of measurement and other records using event group ids by @ChaoPang in #18
- added an option to exclude features and store the cohorts in the meds format by @ChaoPang in #20
- convert prediction_time to timestamp when meds_format is enabled by @ChaoPang in #22
- Combine bin with concept by @ChaoPang in #21
- persist the patient events throughout the generation of the patient sequences for better readability and debugging purposes by @ChaoPang in #24
- updated python-build.yml by @ChaoPang in #25
Full Changelog: v0.0.4...v0.0.5
v0.0.4
put back logging_utils to enable custom logging for spark application…
v0.0.3
What's Changed
- fixed the calculation of the time to event column when the prediction window is unbounded by @ChaoPang in #3
- Feature/add is drug roll up concept by @ChaoPang in #4
- Added units to cehrbert/cehrgpt training data by @ChaoPang in #6
- enabled the continuous job to build the prediction cohorts based on the existing cohorts by @ChaoPang in #5
- added the token_priority module to centralize the token priorities by @ChaoPang in #7
- Handle the categorical measurement records with null value_as_concept_id by @ChaoPang in #8
- set none unit to a N/A string by @ChaoPang in #9
Full Changelog: v0.0.2...v0.0.3
v0.0.2: added time_to_event columns (#2)
* added time_to_event columns * added the python-build.yml