Submission in the Citadel Datathon. Code temporarily dumped without explanation, paper available.
- Files are in
./data/
- data <-
diabetic_data_initial.csv
- discharge_ids <-
discharge.csv
- admission_type_ids <-
admission_type.csv
- admission_source_ids <-
admission_type.csv
- data <-
cmbd_charlson_score
is an aggregate risk score- the other
cmbd_.
variables are logicals representing comorbidity risk - encounter_id is currently a string
- the numeric labeled factors (admission source, discharge, etc.) have new additional description columns