mit-lcp / shakespeare-method Goto Github PK

before executing the notebook 1.0_create_adult_inputs.ipynb I have created a database as follows:
make mimic datadir="/media/joepareti54/Elements/x/finance-2021/AI/NLP/Shakespeare-Booz-Allen/physionet.org/files/mimiciii-demo/1.4/" DBNAME="joepareti54" DBPASS="***”

and then changed the notebook:
from sqlalchemy import create_engine, update conn = psycopg2.connect("dbname=joepareti54 user=postgres \ password=**** options=--search_path=mimiciii"); engine = create_engine('postgresql://joepareti54:****@localhost/mimiciii') cur = conn.cursor(); cur.execute("""SET search_path = mimiciii;""")

because I do not have the variables
POSTGRES_CONNECT = os.environ.get("POSTGRES_CONNECT") POSTGRES_ENGINE = os.environ.get("POSTGRES_ENGINE")

The notebook runs although the data is different than here

The notebook 1.1a_create_transfused_cohort.ipynb however throws an invalid syntax exception, and I believe it does not find any transfusion case:

COPY mimiciii.transfusion_items_dict FROM 'D:\\20180717D_ITEMS_related_to_blood_full.csv' DELIMITER ',' CSV HEADER;

In summary, I need to identify the correct data-set and to fix the above code. More information here.