python -> SQL

[SPTKL]

group by date_permittd year, week, job_type., occ_category on units_net

SELECT Extract('year' FROM date_permittd :: timestamp) AS year, 
       Extract('week' FROM date_permittd :: timestamp) AS week, 
       job_type, 
       occ_category, 
       SUM(units_net :: NUMERIC) 
FROM   devdb_export 
GROUP  BY Extract('year' FROM date_permittd :: timestamp), 
          Extract('week' FROM date_permittd :: timestamp), 
          job_type, 
          occ_category

from sqlalchemy import create_engine
import pandas as pd

conn = create_engine('postgresql://XXX:XXX@XXX:XXX/XXX')
df_permitted = pd.read_sql('''
SELECT Extract('year' FROM date_permittd :: timestamp) AS year, 
       Extract('week' FROM date_permittd :: timestamp) AS week, 
       job_type, 
       occ_category, 
       SUM(units_net :: NUMERIC) 
FROM   devdb_export 
GROUP  BY Extract('year' FROM date_permittd :: timestamp), 
          Extract('week' FROM date_permittd :: timestamp), 
          job_type, 
          occ_category
''', con = conn)

[td928]

resolved in de2fce71d29a07e403b5384f2431f4eabb64f340