def extract_facility_data(inventory_dict):
import stewi
facility_mapping = pd.DataFrame()
# load facility data from stewi output directory, keeping only the facility IDs, and geographic information
inventory_list = list(inventory_dict.keys())
for i in range(len(inventory_dict)):
# define inventory name as inventory type + inventory year (e.g., NEI_2017)
database = inventory_list[i]
year = list(inventory_dict.values())[i]
inventory_name = database + '_' + year
facilities = stewi.getInventoryFacilities(database, year)
facilities = facilities[['FacilityID', 'State', 'County', 'NAICS']]
if len(facilities[facilities.duplicated(subset='FacilityID',
keep=False)]) > 0:
log.info('Duplicate facilities in ' + inventory_name +
' - keeping first listed')
facilities.drop_duplicates(subset='FacilityID',
keep='first',
inplace=True)
facility_mapping = facility_mapping.append(facilities)
# Apply FIPS to facility locations
facility_mapping = apply_county_FIPS(facility_mapping)
return facility_mapping
def get_reported_releases(CASlist):
"""
Retrieves release info from stewi for a list of CAS
:param CASlist: list, a list of CAS in standard CAS format
:return: a pandas DataFrame with records for each release with context and facility information
"""
chem_releases = pd.DataFrame()
for k, v in inventories_of_interest.items():
inv = stewi.getInventory(k, v)
#filter by chems of interest
inv['FlowName'] = inv['FlowName'].apply(lambda x: x.lower())
inv_fl_of_interest = list(chems_stewi_matches[k].values)
inv_fl_of_interest = list(filter(None, inv_fl_of_interest))
inv_fl_of_interest = [x.lower() for x in inv_fl_of_interest]
inv = inv[inv["FlowName"].isin(inv_fl_of_interest)]
inv["Source"] = k
inv["Year"] = v
#Join with facility data to get location
fac = stewi.getInventoryFacilities(k, v)
#Filter by fac in chem_releases
uniq_facs = pd.unique(inv['FacilityID'])
fac = fac[fac["FacilityID"].isin(uniq_facs)]
inv = pd.merge(inv, fac, on=['FacilityID'])
chem_releases = pd.concat([chem_releases, inv], sort=False)
return chem_releases
def extract_facility_data(inventory_dict):
"""
Returns df of facilities from each inventory in inventory_dict,
including FIPS code
:param inventory_dict: a dictionary of inventory types and years (e.g.,
{'NEI':'2017', 'TRI':'2017'})
:return: df
"""
import stewi
facility_mapping = pd.DataFrame()
# load facility data from stewi output directory, keeping only the
# facility IDs, and geographic information
inventory_list = list(inventory_dict.keys())
for i in range(len(inventory_dict)):
# define inventory name as inventory type + inventory year
# (e.g., NEI_2017)
database = inventory_list[i]
year = list(inventory_dict.values())[i]
inventory_name = database + '_' + year
facilities = stewi.getInventoryFacilities(database, year)
facilities = facilities[['FacilityID', 'State', 'County', 'NAICS']]
if len(facilities[facilities.duplicated(subset='FacilityID',
keep=False)]) > 0:
log.debug('Duplicate facilities in %s - keeping first listed',
inventory_name)
facilities.drop_duplicates(subset='FacilityID',
keep='first',
inplace=True)
facility_mapping = facility_mapping.append(facilities)
# Apply FIPS to facility locations
facility_mapping = apply_county_FIPS(facility_mapping)
return facility_mapping
def assign_nonpoint_dqi(args):
'''
Compares facility coverage data between NEI point and Census to estimate
facility coverage in NEI nonpoint
'''
import stewi
import flowsa
nei_facility_list = stewi.getInventoryFacilities('NEI', args['year'])
nei_count = nei_facility_list.groupby('NAICS')['FacilityID'].count()
census = flowsa.getFlowByActivity(flowclass=['Other'],
years=[args['year']],
datasource="Census_CBP")
census = census[census['FlowName'] == 'Number of establishments']
census_count = census.groupby('ActivityProducedBy')['FlowAmount'].sum()
import pandas as pd
import stewi
from os.path import join
from electricitylci.globals import data_dir
from electricitylci.model_config import model_specs
# get egrid facility file from stewi
egrid_facilities = stewi.getInventoryFacilities("eGRID", model_specs.egrid_year)
egrid_facilities.rename(columns={'Plant primary coal/oil/gas/ other fossil fuel category': 'FuelCategory', 'Plant primary fuel': 'PrimaryFuel', 'eGRID subregion acronym': 'Subregion', 'NERC region acronym': 'NERC'}, inplace=True)
# Remove NERC from original egrid output in stewi because there are mismatches in the original data with more than 1 NERC per egrid subregion
egrid_facilities = egrid_facilities.drop(columns='NERC')
# Bring in eGRID subregion-NERC mapping
egrid_nerc = pd.read_csv(join(data_dir, 'egrid_subregion_to_NERC.csv'), low_memory=False)
egrid_facilities = pd.merge(egrid_facilities, egrid_nerc, on='Subregion', how='left')
len(egrid_facilities)
# 2016:9709
egrid_subregions = list(pd.unique(egrid_facilities['Subregion']))
# Remove nan if present
egrid_subregions = [x for x in egrid_subregions if str(x) != 'nan']
len(egrid_subregions)
# 2016: 26
# egrid_subregions = ['AZNM']
egrid_primary_fuel_categories = sorted(pd.unique(egrid_facilities['FuelCategory'].dropna()))
# correspondence between fuel category and percent_gen
fuel_cat_to_per_gen = {'BIOMASS': 'Plant biomass generation percent (resource mix)',
import pandas as pd
import stewi
from os.path import join
from electricitylci.globals import data_dir
from electricitylci.model_config import (
egrid_year, min_plant_percent_generation_from_primary_fuel_category)
#get egrid facility file from stewi
egrid_facilities = stewi.getInventoryFacilities("eGRID", egrid_year)
egrid_facilities.rename(columns={
'Plant primary coal/oil/gas/ other fossil fuel category':
'FuelCategory',
'Plant primary fuel':
'PrimaryFuel',
'eGRID subregion acronym':
'Subregion',
'NERC region acronym':
'NERC'
},
inplace=True)
#Remove NERC from original egrid output in stewi because there are mismatches in the original data with more than 1 NERC per egrid subregion
egrid_facilities = egrid_facilities.drop(columns='NERC')
#Bring in eGRID subregion-NERC mapping
egrid_nerc = pd.read_csv(join(data_dir, 'egrid_subregion_to_NERC.csv'))
egrid_facilities = pd.merge(egrid_facilities,
egrid_nerc,
on='Subregion',
how='left')
len(egrid_facilities)
inventory='TRI'
year = '2016'
#Get one of these inventory
tri2016 = stewi.getInventory(inventory,year)
#See first 50
tri2016.head(50)
#Look at all the unique flows in this inventory
tri2016flows = stewi.getInventoryFlows(inventory,year)
#See first 50
tri2016flows.head(50)
#Look at all the unique facilities in this inventory
tri2016facilities = stewi.getInventoryFacilities(inventory,year)
#See first 50
tri2016facilities.head(50)
#Now combine with some inventories in another inventory based on facilities
#Enter inventories that you would like to combine in the "Inventory_acryonym":"year" format enclosed in "{}"
inventories_to_get = {"TRI":"2016","NEI":"2016","RCRAInfo":"2015","eGRID":"2016"}
base_inventory = inventory
combinedinventories = stewicombo.combineInventoriesforFacilitiesinOneInventory(base_inventory, inventories_to_get)
#See first 50
combinedinventories.head(50)
#See a summary of the combined inventories by facility and flow
pivotofinventories = stewicombo.pivotCombinedInventories(combinedinventories)
pivotofinventories.head(200)